:)\J

A STUDY •,- ~':,_\~~~

BY

T. R. DAWSON and W. HUNDSDORFER.

EDITED BY

· ALAIN C. WHITE.

LEEDS:
 1

TO
M. W. VJ.

-----.
 , '

7. Legal En Passant Keys .. · ... Nos. 52-97

(a) Theme Pawn absent from Passage Square, to prevent
(1) Imaginary Check on the Side Square.
·/i.) Black Retrograde Stalemate Nos. 52-62
(ii.) White Ret_rograde Stalemate No. 63
TABLE OF CONTENTS. (iii.) The Imaginary· Check deferred No. 64
(2) Horizontal and Diagonal Line Obstructions.
I. ILLEG_.\.L POSITIOXS.
(i.) Black Retrograde Stalemate.
r. Retrograde Stalemate Nos. r-2 Checking themes
2. Excessive Pa,vn Cap~ures No. 3 (ii.) White Retrogra~e Stalemate.
,,
3. Illegally Situated Pieces Ncs. 4-5. Checking themes )so. 68
4. The Correction of Illegalities Indirect Checking themes Nos. 69
Kos. 6-9
(a) Partial _.\.nalysis Non-checking themes Nos. 70-71
Kos. IO-II
(3) Yertical Line Obstructions.
(b) Turn to )Iove Nos. r2-r5
Checking themes Nos. 72-73
II. LEG1L POSITIOXS. Iudirect Checking themes No. 74
A. Analysis of Past Play "·ithout future Purpose. Xon-checking themes Nos. 75-76
r. Incidental Analysis No. 16 (4) Square Blocking.
2. Obtrusive )fen (i.) Non-checking themes.
Kos. 17-18
3. _.\.nalysis as the Sole Aim Double Retrograde Stalemate Xos. 77-79
Nos. 19-25
White Retrograde Stalemate Nos. 80-81
(a) Turn to )Iove .. Nos. 22-25
(ii.) Checking themes.
B. Analysis of Past Play "-:ith a future Purpose.
Black Retrograde Stalemate Xo. 82
4. Disproof of )Iiscellaneous Tries No. 26 White Retrograde Stalemate Ko. 83
5. Legal Retractors Nos. 27-31 (b) Theme Pawn present on Original Square, to allow
6. Illegal Castling .. Nos. 32-51 ( r) I,ine Covering.
(a) Retrograde Stalemate-:- (i.) Black Retrograde Stalemate No. 84
(!) Present Nos. 3 2 -34 (ii.) White Retrograde Stalemate
(2) Past .. ,. Nos. 35-37 Covering the Black King · Nos. 85-90
(b) Castling King Moved j CQ\·ering the ·white King Nos. 91-95
Nos. 38'-:-4!;.
(c) An~ l.'northodox Expedient Nos. 96-97
(c) Castling Rook Moved Kos. 46-51 8. Partial Analysis Nos. 98-100
4
5
 '/No. L No. L

T. KING--:PARKS. r Se6.

C.P ..4. Magazine, 1908.

. The average book of chess problems contains diagrams
· · in which something is expected to happen, usua~y a ma~e
BLACK. or a self-mate. In this book, too, the problems will contam
mates or self-mates, save in a very few cases ; ~ut the
solutions will be simple, often elementary, _and the m!e.rest
will centre on what has happened in reachm_g the_ pos1t10ns
diagrammed, supposing each to hav~ . ansen ~n actual
play. · We will begin with some positions wh~c~ could
never have occurred in actual play. Such pos1t10ns w_e
may call illegal. The present group is illegal, because it
is White's turn to move, and Black could not have played
last.
Xo. IA. No. IB.
L. :N". de Jong. T. R. Dawson.
Op de Hoogte, 1912. Birmingham Post, 14 July, 191 r.

WHITE.

Mate in two.

WHITE. WHITE.
j- Mate in two. Mate in two.

I QaI.

. 6 7
 , .

· No. 2.

H. HJORTH. No. 2.
Skakbladet, Jan., 19rr.
I P=S, Pg6 ; 2 Kh6. _
BLACK
We use with forethought the word illegal to define any
\
condition which could not arise in actual play. The word
impossible is often used in the same sense, but it is not
satisfactory, and we shall not use it. There is no such
thing as an impossible position, provided you have enough
chess-men in your box to draw on. The word always
provoked Sam Loyd. "Impossible? " he would say,
" you say these men could not have got into such a posi-
tion ! Why, they are in that position ; I put them there
myself ! " To this no answer can be made. But legality
is another matter, definitely demonstrable or the reverse,
as we shall ·duly see. Legality is often disparaged (see
p. 4r). Its requirement is avowedly a convention. It
does not much interest the solver. Judges, editors, com-
posers bow to its dictates to humour their consciences
rather than to gratify their taste. We only know of one
article dealing with retrograde stalemate, to use the technical
term for the legal impossibility of Black having moved
last. This is L. N. de Jong's Onmogelijke Standen in Op de
Hoogte, from which No. IA is drawn. ·
WHITE. Of course any accidental retrograde stalemate can ·be
, legalised by the judicious addition of a harmless Black
Mate in three. \ Pawn. We remember our surprise at finding No. 2 re-
produced by a careful editor with an added White Pawn
at az and a Black Pawn at a3. Their remoteness from
the field of action drew attention quaintly to the grave
illegality they were intended to correct.
I-

-- ----4-. - ·-
 l.

No. 3. No. 3.
B. \V. LA MoTHB. r Qd5.·

Wilkesbarre Record Ty., 1889. The first question the investigator -of legalities looks
into is whether the Pawn position could have occurred in
BLACK. actual play, whether the captures necessary to bring the
Pawns into their places can be accounted for by the men
captured from the enemy. If, in No. 3, we find eight
captures necessary and only six men lost by Black, we
at once convict the position of illegality. Sometimes
this is rather a puzzling matter to determine, as in No.
173 of Les Tours de Fqrce./C Composers use the principle
of excessive Pawn captures as a solving tourney catch.
Forty solvers failed to see that Black's Q R could not
have· been captured by a White Pawn in No.' 3A. Other
composers use the principle foi: bizarre effects, as in No. 3n.
No. 3A. :No. 3B.
E. Eginton. W. H. K. Pollock.
I Birmingham Post, 28 May, 1912. Balti11101'e News, 29 June, 1890.
• BLACK. .
/
I I.?

WRITE.

Mate in two.
 No. 4. .,
0. EwETZKY.
No. 4.
Schachmatnoe Zadatsch, 1908.
1 Qb3.
BLACK.
A third type of illegality arises from placing pieces on
squares to which they could not have arrived in play.
Take the Ph4 in No. 4. In actual play only White's
KR P, KS P and KB P can ever reach h4. As these
P's are· all on the board in No. 4, we suppose the illegality
of the'Ph4 to be an oversight, readily corrected by moving
the whole body of men one square to the left. It is
important to understand exactly wherein this case of
illegality differs from those in the last group. As we
advance into increasingly complicated tangles of the
chess-mert, the presence of elementary illegalities will have
to be taken for granted in most instances, owing to sheer
lack of space. Next to the Intruding Pawn, the most
frequent violator of the legalities is the poor Bishop,' who
has a way of getting in behind Pawns (No. 5A) on squares
he could never have re'.'l,ched even by promotion. The
inartistic, but entirely legal, case of the Obtrusive Bishop
will be discussed on pages 39-41.
For sh..9..:cLw.e_ma~alL:y:._si~ed piece an
imaginary piece. Sometimes, though only by ··a conscious
WHITE. effort or'f!:ieaiiaf"ytic composer, several pieces get bottled
up together in a general illegal mix-up. We refer to such
:'.I.fate in two. pieces as occupying an imaginary district. These further
resolve themselves on a few very interesting occasions
into a state of perpetual retrogr9.sion, which we will not
explain until we first meet with 1t in No. 64.
i

12
 No. 5.
No. 5.
I Bez.
A. \VATERHOUSE. .
. '
" Here is a conventional tour of White Knight mates
with nothing at first sight_ illegal about the setting, just
Birmingham Post, 30 June, r9r4.
the sort of thing to play havoc with tourney solvers, for
how could the Black Rook ever have got to gz? It must
BLACK.
ha.ve corn.e. around behind the White Pawns, passing over \
gr, and ori. that square giving an illegal or imaginary
check to the White King; We call ~lreck'tliat co'ulu-
nofi>ccliHii''actual play an imaginary check. The Germans ·
give it the more scientific name of an irreal .che.£k., No. .
5B is illegal for the same reason, that tii"e White King in
crossing the sixth line must have encountered an
imaginary or irreal check from one of the Black Pawns.
Compare the position of the White King in this last with
No. IB on the assumption there that Black's last move
was Pc7 or e7 X d6. See also No. 13 (page 3 r).
:N°o. 5A. Xo. 5B.
i
C. Guarini. S. Hermann:
Western.Daily Nlenury Ty., rgog. Gazette Times Ty., rgrr-12.
BLACK. BLACK.
•~===""' ::::;;;;;;;,=::::;;;;,:t

WHITE.

l\Iate in two.

Mate in two. ::Vfate in three.

 No. 6.

J. PALUZIE.

Leeds Mercury, 10 Dec., 1898.

No. 6.
A considerable body of puzzles, no one can seriously
call them problems, have been composed on the idea of
correcting illegalities. In other words an illegal position
is set up, and the solver has to discover the illegality and
correct it before any solution will operate. In some
cases, as in the puzzles of Group 7, where both Kings are.
standing in check, the fact that an illegality exists is
obvious from the positions, but it is not demonstrable in
character. In other puzzles, like No. 6, the illegality is
concealed and its character is supposedly determinable.
This statement requires the qualification we have made.
For it is impossible to prove the c;ause of an illegality in
any case ; the best we can do is to take evidence as to the
most plausible or immediate cause that is present. In
No. 6, Black's King's Bishop is clearly illegally placed.
He cannot be a promoted Pawn, for a very little study
will show that no Black Pawn can have been promoted
on a Black square. Such a promotion would involve what
we have called excessive capturing of the White men.
Consequently the author claims that this Bishop either
WHITE. does not belong on the board at all, and should be removed,
or else it belongs at £8, whence it could not liave moved
Mate or Self-mate in one. out, and should be put back there. In either case White
can mate by I BX R, or compel self-mate by I Qb8.
if-
No. 7. . No. 7:
ll
G. Hmm. In each position of this group both Kings are in check,
and we are obviously confronted with something decidedly
B.C.M., Dec., 1893. illegal. In No. 7 Black has just pfayed a. Pawn from gz,
taking a White piece on hr and claiming the Black Knight.
BLACK. But as he thereby discovers check im his own King, White .
calls upon him to retract his move and take the piece hr
legally, whereupon White self-mates by BX P +. In
No. JA Black has just played Pdz X S=S. Replace the
move and ~xact the King move penalty. Then Qh8 mate.
In No. 7B Black in his eagerness to mate has played Queen.
from hz, g3, 4, 5 or 6, quite forgetting the 'liVhite Rook
check. Replace the Black Queen and demand a legal
Queen move, and according as the check is covered at
c7, 6, 5 or 4, then Sd6, PX P, Rd4 or Rd5 mates-the
Black Queen interfering with four of her owh pieces. ·

:No. 7A. Xo. 7B.

B. G. Laws. T. R. Dawson.
Leeds JJercury, 12 Dec., 1891. Original.
BLACK. BLACK.

WRITE.

Self-mate in one.

Mate in on~. ::\fate in one. ·

18 19

------
 t·

•s161
'smu.Nl-;d 'R3.'l'l1W a.NV
·a.-,v•1 aoo,Y.W'lll 'sI : saaa'I

t
I

r~
.I ·:,rJ,IHA\ NIV'1

.&a: as:J.ra:,r
 No. 8 ..

T. B. ROWLAND.
No. 8.
Leeds Mercury, 12 Dec., l:896.
The demonstration of the precise nature of the illegality
BLACK. to be corrected is much more plausible in No. 7 than in
Nos. 7A-B, because in the latter other suppositions would
explain matters just as well. The reason we accept the
composers' explanations is because these permit the neces- '
sary mates in one and not because they appeal to our
analytic powers. Therefore we say that No. 7 is a much
sounder study than the other two, meaning that it is more
plausible. No. 8 is likewise very plausible. The position
of the White Pawns makes it appear that they have cap-
tured thirteen men, whereas only twelve are missing.
To account for this we must consider Black's previous
move. The Rook could not have moved, since the Bishop
could not have given check on g8, Nor could the King
have moved. It must then be the Pawn on b5 which
could only come from b7, showing that the White Pawn
has played from c5, with the intention of capturing en
passant. But this would be illegal. He must therefore
retract his false move and move his King as penalty.
Fortunately the White King's only move (r Kb6) gives
he desired mate in two.
We· shall be constantly writing retrograde moves.
WHI'.rE.
The most convenient 'Yay is to give such movec, in their
direct sense, e.g., Pc5-b6 in No. 8, leaving understood
Mate in two.
tVe fact that a tetraction is being made. The importance
of a clear conception of these simple technicalities cannot
be too forcibly impressed upon the reader who wishes to
follow with ease our subsequent classification.

20
 No. 9. No. 9.
In. writing 'retrograde moves special attention is
A. BATORI.
demanded by the fact that an opposing piece may reappear
on the square vacated. Thus, in No. JA, where the Black
1lfagyar Sakkujsag, Nov., rgn. P at dz supposedly captured a White S at er, we say the
retrograde move was Pdzx S, and hereafter we omit the
BLACK.
word ~~ This process is the converse of capturing;
and for 1t we may employ the term U.!!:fapturing. To
uncapture a man is to replace a move which·· involved a
capture. ·
Numerous puzzles hinge on showing that Black has
Castled illegally. and m~st {tnccistle!- whereupon enforcing_
the penalty White readily mates. "No. 9A-B dispute the
claim to priority for this idea. No. 9 is the best rendering.
The Black K moved earlier in the game to permit \Vhite's
. K]?P, to promote at e8 or f8, so that Black's having just
r-·-Castled was clearly illegal.
~~/::, Xo. 9A. No. 9B.
A. Konig. G. T. Robertson.
Svetozor, 1882. · Brentano's, July, 1882.
BLACK.

1\fate in two.

WHITE.

llfate in one. Mate in two.

22 23

----------·-- --- ·-. -

~
 No. 10. ·

J. _P ALUZIE.
No. 10.
Ruy Lopez, Oct., 1896.
I. The puzzles we are considering in these groups (6 to r5)
BLACK. I,,, might have been sub-divided into the same three classes
as the examples "of illegality given in Groups I to 5. No.
I2, for instance, depends for its solution on proof of retro-
grade stalemate; No: 8 on proof of excessive Pawn cap-
tures; No. 6 on proof of an illegally' situated piece. We
have not adhered to such a sub-division, because the
effects produced merge almost imperceptibly into one
another. There is practically no difference between No.
IIA (excessive Pawn captures) and No. IIB (imaginary
Pawn), because of the great similarity in the method
adopted to correct the illegality. Consequently each
group is intended to show some method of correcting an
illegality, rather than a particular illegality corrected in
different ways. No. ~o contains an illegally situated
Black King (compare the White King of No. SB). Our
recollection of No. 6 suggests that we put the Black King
back where he belongs. But where is that? May .he
not stand on a lot of squares, a8, c5, fa, g6, hr or even
others ? Surely. Each of these is a legal square, and
for each we shall find a corresponding, often very pretty,
mate in one. This type of correction of illegalities may
well be called Partial Analysis. We cannot determine
accurately how to correct a position ; but we can find out
l\Iate in one. by analysis what is most plausibly wrong, and each way.
of righting it will yield a distinct solution. (See also
No. JB.)
j-

---- ~-.
 No. 11.

Sometimes an illegal situation can be straightened out

~ L'Illuslration Universelle, 1868. by removing a man coqipletely from the board. In No.
II the Black Pawns have made eight captures ; only seven
BLACK. White men .~re missing ; remove therefore any piece or
Pawn of \Vlute (except of course the King) and mate will
follow forthwith. In Ro. IIA it is Black's force that is
I.
~uperabundant. \Vhite's seven Pawn captures are ;
imperfectly b~lanced by only six missing Black men. If I
you correct this by the removal of any one of the remaining
Black men, the desired mate in one will be revealed. No.
rm goes a step further, presenting nine Black Pawns.
As this passes our belief we remove one of tbe Pawns and
mate in one results.
Xo. IIA. No. IIB.
G. Chocolous.-
J. Keeble.
Oesterreichische Schaclzzeilung,
).fay, 1873. Jamaica Gleaner, 27 Dec., 1890.

WHITE.

1, c d.'-I
'·
~(

-:-, ·l 4
-.
'l -
:\fate in one.

--
J r!
r
'l ·-- V•
WHITE.
17 --,4
lt

~-
- '..i
).fate in one. Mate in one.

- •,
'1 (;f':,
!1
''
C'->i 4
~ J? ~£
~ r t&

---·-·
 No. 12.

\V. A. SHINKMAN.
No. 12.
Chess Journal, 15 Nov., 1870.
This position is quite as obviously a case of retrograde
BLACK.
stalemate as the problems in Group 1, since no legal last
move can be found for Black. We have, however, now
advanced to the ability of correcting the illegalities we
meet; so we ask ourselves, "What is wrong with No.
12?" Since Black cannot have moved last, White must
have done so. Therefore it is now Black's turn to move,
and whatever move he makes (Kx S or Pf6) White mates
in one. This is no longer a case of Partial Analysis, since
the choice of moves occurs as a result of different possi-
bilities springing from the initial position, like variations,
instead of supposedly leading up. to it in different ways
from earlier conditions. By analysis we mean the explana-
tion only of moves or events which are supposed to have
taken place prior to reaching the positions given on our
diagrams. In other words we determine by Retrograde
Analysis whose turn it is to move, and where we find that
itis Black's turn to move we let him do so before proceeding
to try and mate him. We are evidently still in the midst
of puzzles, on the border line of eccentric chess ; but
we are approaching much nearer to the field of Analysis
unmixed· with imaginative speculations or guesses. Our
meaning will be made clearer by turning from No. 12 t&-
No. 14. If the ·condition of the latter is written " Mate J.
:Mate in one. in two," we are confronted by a puzzle; if it be written .,I
" The legal player mates in two " we have reached the
domain of accurate analysis. ·
I

rl "
'

. 29
 No. 13 ..
No. 13.
B. G. LAWS.
The same process will serve ·us here as in the case of
No. IZ. \Ve recognise in each position of the group a
Jamaica Gleaner, 31 Dec., 1892.
retrograde stalemate condition ; so we let Black move
it being his turn, and then give mate in two accordingly'.
BLACK.
The mate in No. I3 after Black has moved P x P and that
in No. I3A after Ph6, have pretty points. No. I3~ presents
a_ clever ~hameleon Pawn echo. . At first sight such posi-
tions remmd us of three - move direct mates with the keys
~ut 3:way. They are indeed little else, provided we bear
m mmd the retrograde stalemate feature. In No. IJ it
will take a moment's thought to see why Black could not
have just played Pg7xf6. Reference to No. 5B reminds
us that such a retraction would render illegal the White
K's station on d7. ·

Xo. 13A. Xo. 13B.

H. D'O. Bernard.
R. Steinweg.
Chess Amateur, Dec., 1914.
r;:=~==BLACK.

WHITE. [:
1:Iate in two.
\
i
I
Mate in two. :1\fate in two.
I
 I·

No. 14.

T. R. DAWSON.

Bolton Football Field, 24 Dec., rgro. No. 14.

(/J
There rp.ay be readers who consider us severe in that
we classify these Turn-to-move positions among the
puzzles involving the correction of illegalities, and not
among the orthodox analytics to which we shall presently
be coming.· Our explanation springs from the definition
we have given of a puzzle, that it is a position whose
solution is at best only plausible, not legally provable if
the solver cares to be captious. There is no fundamental
difference between No. r and No. rz and No. r4. Yet
of No. r we say that it is a straightforward two-mover in
which the composer has hit upon a retrograde stalemate,
that is upon an illegality of position. Then in No. rz we
say that the retrograde stalemate can be legalised by letting
Black move first at will, after whjch we shall be ready to
mate in one. Finally in No. r4 we say that the retrograde
stalemate can be legalised once more by letting Black ·
move first ; but as it is his move the mate shall be given
by him (r Qx P + ). Here are three different explanations
of three positions alike in their condition of retrograde
stalemate, and we choose the most plausible to suit our-
selves. This is not analysis, it is puzzledom. . As a puzzle
WHITE. No. r4 offers some pretty incidents, especially the reasons
why a possible last move for Black could not have been
Mate in two. Pb7 X a6, or Pe7 X d6, or Pg3 x hz. The position of No.
14 can be arrived at in actual play in 33 moves, also not
an easy matter fo demonstrate.
 No. 15 ..
No. 15.
J. JESPERSEN.
The Turn-to-move puzzle,·with the mate administered
by Black, was a favourite with the late J. Jespersen~
Wochenschach, 31 Mar., 1907.
That ingenious composer, among his phenomenal output
of 3,000 problems, has left us a little series of this nature
from which the present group is· selected. In No. r5 th~
solver is supposed to try r Pb7 +, only to find this illegal,
because Black could not have moved last. The real
solution is for Black to play r Rh8 +. A last move for
Black of Phz=B is imaginary, as then the Pawn must
have come from c7, capturing on the way White's Queen's
Bishop, whose presence away from er is not explainable.
Nos. r5A-B are clever twins, the former solved by BSg5
mate, the latter by WPd5 mate, according to the presence
or absence of retrograde stalemate in the positions.

Xo. 15A. No. 15B.

J. Jespersen. J. Jespersen.
Tidskrift, 24 Dec., 1898. Tidskrift, 24 Dec., 1898 .
.</~ BLACK. /o
==·

Mate in two.

WHITE. /_,/
~:L
Mate in one. Mate in one.

· 34
 No. 16.

No. 16..
l\Ianclzester Times, 19u.

BI,ACK.
I Kd4.

, The groups in Part I. of this· book have given us an

idea of the elementary types of illegality which are to
be found in chess positions. From now on all our positions
will be legal ones, and the illegalities will occur only in
our search for proof as to whether or not different hypo-
thetical explanations of the details of each position are
sound. In P_~.,__illegaj,itieLfig11r~<:l _in every poshio11 ;
in Part II. they will figui:e onlyjn the retrograde_ tries'.-
j,J\
rn Part r. -analysis revealed that the--pieces-could not have 1
reached their positions in actual play, and conjecture of
a mor~ or less plausible kind sometimes suggested correc-
tions by which the positions could best be made legal.
In Part II. analysis will reveal not only that each diagram
is a legal one, but it will furnish some evidence as to the
character of the past play, usually as to the last move of
Black, but often as to some earlier and indeterminate play. .
Retrograde knowledge which such analysis furnishes can
be used for several purposes ; to legalise the retraction
of moves, to prevent one's opponent from Castling, or to
WHITE.
permit an En Passant key move. There are cases where
analysis will give us information of a curious, but quite
l\fate in two. purposeless, nature. For instance, in No. r6, Black can
without difficulty be shown to have Castled on his last
move. The problem differs from No. 9 in being quite
legal; but having obtained our retrograde information
we can make no use of it. ·
i

37
 No. 17:
No. 17.. _,.
I Pb4, Pd2 ; 2 Qgz.
H. w. BARRY.
Retrograde analysis in positions like No. 16 may be
Checkmate, Jan., 1903, characterised as ornamental, or accidental, or better still
as incidental. The upholder of ar~istic laws uses incidental
BLACK. analysis to detect the presence of promoted pieces in ·
problems, to_ which he refers as Obtrusive· M:en. The
"artistic" argument against the Obtrusive Bishop, who
must not be confounded with the Imaginary Bishop of
No. 5A, was discussed on p. 449 o~ the Loyd BQok. It
concerns only the tourney Judge, and the scrupulous editor
and composer ; and in no way affects the retrograde student,
in whose complicated positions the artifice of promoted
men is a sine qua non. ·

No. 17A. No. 17B. ,

H. H. Davis. T. R; Dawson.
2 Pr., Bristol 1Vlercttry, . Wes/en imd Daheim, 2 May,
6 Nov., 1886. 1915.
BLACK. BLACK,
r;:=:::,,~=

WHITE.

Mate in three.

j- Mate in two. Mate in three.

-
I Qf8. l Bf7.

39
 No. 18. · ·

J. CRAKE. No. 18.

B.C.M., April, I889.
r Qc5, SxQ; 2 Sf5+.
BLACK.
The persiste;nt obtrusive Bishop finds his way into
many excellent problems, and No. r7A showed him brazenly
,vinning a prize. His cousins, the Obtrusive Rook and
Knight, are much rarer fellows. No. r7B has been com-
posed to exemplify the latter (see No. 42), and No. r8 may
possibly be the only unconsci_ous_ example of ~he forme_r;. :C...tJ..o
In both cases the ,yord obtrusive is almost a imsnomer; as 1 ~,.,:, ,.icf:~
the reasons why, m No. r7B, the Se8 must have been · ,
promoted and, in No. r8, Black's K R couJd never have t
· left home, are anythin_g but obvious. Those who belittle) , . ~-
i LOA_ c'i CJ::: the...bugbear o~ Obtrusive. Me~ often go further and arg~t-J'-!.d,nw (
-:f , . . that the quest10n of legality is an. absurd one. Problems, f\
C/.'v~ they say, have emancipated themselves from the game, !\.
, and the principle of the probability of problem positions f\
j .i:,:f;c.:J\ hasbe~~ quit~-32!!!:ft.I~...
~ \ Whr should we not thro~ off ji
'/ ,, also the shackles of this convent10n of a necessary legality? J.
Loyd often disregarded it with a twinkle in his eye, and jh
many moderns do so when the need arises, finding their l[
justification in t~e i1;teresting. points of view suggested . I!
by A. Gehlert, m his revolut10nary pamphlet of r903, •;
On the Nature of Chess Problems. As believers in the b
·wHITE.
varied and profound beauties of retrograde analysis, we ii·
Mate in three.
must decline to accept Gehlert's views unr~servedly, II ·
though we may add that were. the law of legality to be ·11
allowed to lapse, the analytical problem would still be a Iii
possibility by simply repostulating the law as a condition 11:
in t}1e enunciation of each J>Osition. II
.·11 ·
·,;
!
'
 No. 19.
No. 19.
l Ke4-d5, Pe5 X P e.p. +. ; 2 Pf7-f5.
F. C. COLLINS.
Incidental retrograde analysis can be given an increased
r Pr., Chess Monthly, April; c88o. importance by being made the composer's sole aim, and
Groups Nos. 19-25 are selected from the· numerous posi-
BI.ACK. tions in which nothing is expected of the solver except
that he discover what the last few moves have been.
We have already had opportunity to notice how frequently
Castling and Captures en passant figure in retrograde
analysis. This is due not only to the unique character
of such moves, but also to the fact that it is in relation
. to these two subjects that a knowledge of the past best
helps us in the future. In consequence these moves
crop up in incidental analysis as well: Castling in No. 16;
En J!assant captures in the present group.

Xo. r9A. Ko. r9B.

S. Loyd. L. Bertrand.
Chess Association Congress,
r8 r. Echiquier Francais, July, 1906.
9
BI.ACK. == ::;::;:;,,:,=:;;;;;BI,;;;;;,A,,,,,CK=;.;;;;;;;,==;,~I

WHITJ;;.

It is White's move.
\Vb.at were the last moves?

':,,

42.
 f

· No. 20.
I
N. HoEG.
I
No. 20.
Original. . I Rd8-d7 +, Pd7-d6 ; 2 Pf5 X P e.p. +, Pe7-e5 ;
3 Pf4-f5 +, Kd6-c7 ; 4 Pb5 X P e.p. +, Pc7-,--c5 ;
5 Pb4-b5 +, Ke6-d6 ; 6 Pg5 X P e.p.f, Pf7-f5 ;
· 7 Pg4-g5 +. .
Here is another remarkable en passant study, the
capture being required three times in the only legal series
of antecedent moves. This is the first long repetition of
retrograde checks we have come across ; and as, on p. 23,
we spoke of uncapturing moves, so in No. 20 we may call
White's retrograde steps imchecking moves, though we
indicate them with the conventional checking sign. In
other words we now understand, from the law of alternate
motion in direct play, how, in retrograde play, as soon
as one player retracts his King into check, the opponent
must reply by retracting a move abolishing the check.
An imaginary or irreal check (see p. 15) results when there
is no way of abolishing the check.
Again, from the law of direct chess that one must not
put his King into check,-we have the retrograde law that
one must not retract a move which leaves· the opposing
King in check, as that also would constitute an imaginary
state. These irreal checks affect the question .of pinning.

J
A piece which would be pinned in direct chess can be moved
. in retrograde chess, provided one's King is not thereby
·what were the last moves ? left in imaginary check. The ~retrogr...ade ·eCJE.!valent
of a pin occurs when a man covers the action of another
or the same colour on the opposing King. What would be.
. a direct batter~ ordinary chess becomes a retrograde
pin when moves are being retracted. (See-p. ng.)
.....
• II ·
1:... -•«•·-· ·-----·· ··•·-· ··•--····••O,•s<·· ·-· --·-~•

i.

44 · 45
 .,

No. 21.
No. 21.
The last moves must have been: r Qhr-a8 +, Ka8~
T. R. DAWSON. a7 ; 2 Rb7--'b8 +, Kb8-a8 ; 3 Rc7-b7 +, Kb7-b8 ;
4 Rc6-c7 +, Kc7-b7 ; 5 Rd6-c6 +, Kc6-9 ; 6 Rd5-.
No. 1,000, Gazette Tiines, 13 July, 1913. d6 +, Kd6-c6 ; 7 Re5-d5 +, Kd5-d6 ; 8. Re4-e5 +,
Ke5-d5 ; 9 Rf4-e4 +, K~-e5 ; IO Rf3-J4 +, Kf4-
BLACK.
e4 ; II Rg3-f3 +, Kf3-f4 ; I2 Rg2-g3 +, Kfa-£3 ;
r3 Rgr-g2 +, Ke2-fa ; r5 Bg2-fr +, Kd3-e2 ; r6 Sc4
-b2 +, Kc2-d3. Besides these sixteen squares the
Black King ha4 to reach c2 via dr, er, fr, and the twentieth
square is e8 !
In writing a series of retrograde ·moves we place the
last player first, just as in a direct series we plaee the next
player first. When, as in No. zr, White has played last,
the White moves come first reminding us of direct chess.
But in most retrograde problems Black has played last,
and his moves come first, which is entirely logical but a
trifle confusing till one is familiar with the technique.
Xo. 21A, Xo. 2IB.
)l"_ Hoeg. N. Hoeg.
Tidskrift, Mar., 1904. Tidskrift, :\far., 1914.
BLACK. \1..,
t'i. ~- 1/,"-"11~1/,½ =======

WHITE.

The Black King never moved twice to any square.

Demonstrate twenty squares he must have occupied.

---------· ---- ·-. -.

 No. 22.

T. R. DAWSON. No. 22.

Pittsbu1-g Leader, 25 May, 1913. In most retrograde problems it is assumed that it is

Black's turn to retract, just as in direct problems it is
BLACK. usually supposed to be \Vhite's turn to move. In Nos.
20-21 we realised that it was White's turn to retract first,
from Black's standing in check. In Nos. ·22-25 the Turnr
to-).Iove analytic is given its broadest form, and we accept
it as a logical study since it is freed from those imaginative
conditions of mate givip.g which classed Nos. 12-15 with
the puzzles. ·
~o. 22, as one of our first complex analytics, deserves
a close study. Black's K B must have been captured at
g3 ; so that the P at g6 has been moved a considerable
time, his move· liberating the B and boxing up the Sh8.
White's KB was captured on c6, and the other five missing
Wbite men fell in the path of the Q R P now on fa. The
P at e3 moved up_ after ;the Bl. P had reached fz and finished
its captures. Heh"•tefath White Rook~ came out on their
own files, before which Wbitef'iilt':t§t '.Rave-played Pa4 on
the left and PX B on the right. White's KB P could not
recently have stood on fz, owing to the Bl. P, or on f3, as
. this was required for the passage of White's KB~ Wbite
then has had only five recently available moves : P from
a4 to a5-a6-a7, and P from f4 to f5-f6. Previous to
these he must have moved his K, and Black's last moves
must have blocked the K up in the shortest way ; thus,
retract, 1 BaZ:-b1, Pa6-a7 ; z Sb3-<!1, Pa5-a6 ;
Who played last ? 3 Qb1-d1, Pa4-a5 ; 4 Rdr-d2, Pf5-f6; 5. Rcr-d1,
Pf4-f5; 6 Rd2-e2, Ke2-fr; 7 Rd1-d2+, Kfx-e2.
As,. only such a sequence will free the White King in time,
the deduction is incontesi;able that Black played last to
¥1e position diagrammed.

, , r • \ -\ -
Q I ' ,r, Q / ~
l •.✓,.,. l"j? 1 n. -6l 'I:.• 1l"<--..//...J-. /, ') r
',t:' ~
_,

/"""'°,( \,'("\v"'<"' -- (}.tp ...;.,

...,vt,\,.).. '\v,JY\ f" \ J

\f f ../ .\ \1 clr\M,t W\.,-t,'t'wie,aiv~NJ~ J~ ! ~ 1 ·

l\l..",\ t;.,Y.,-< r," ,~.v t«} f"'° ( 49 D .•
\ \, I
 ;

No. 23. No; 23.

- '

T. R. DAWSON.
In No. 22 we saw that any retrograde series of inoves
· dissimilar to th'e author's solution would have boxed up
the White K and with him all his army in a condition
Gazette Times, 24 Dec., 19u. analogous to what, in Nos. I and 13, we called retrograde
stalemate the only difference being that heretofore in our
BLACK.
examples' the stalemate was present in the diagran:s,
while in No. 22 the stalemate would have cropped up in
an imaginary past were incorrect retrac_tions 1:11ade. So i_n
direct chess a stalemate may appear m a diagram, or it
may crop up in the tries to the solution of an Indian or
other kind of problem.
Avoiding retrograde stalemate in No. 23, we _find that
only White can have moved last, the moves having been :
r Kc8 X B, Bb8-c7 + ; 2 Pg4-g5, S~6-f8 ; 3 Bc7-d8,
Sd8-e6+.
. ~o. 23A.
-- . ··
· , f--:JO ,1
: v\ · -,,. w
;
v
~T
No. 23B. ft
i,· i.r8 ·1)'
'.•'.,j~ 1,r1.f".,..,,,· ...,,/..
t,/',.1. UJ,.,, ·.
T. ·R. Dawson. : J. J esperse . ~-/-, , \.::--( f VI . ~ "l.. :.
Wiener Schachzeitung, 1912.. \ lViener Schachzeitung, Fieb., 1901. (f ·,
BLACK. -·I l'\ • BLACK. 'r'
/,<... ''
J
\ '

WHITE.

Demonstrate that Black has given two checks in his

three last moves.

50

•
 T. R. DAWSON:

Gazette Times, 24 Dec., r9rr. No. 24.

. BLACK.
It is not usu~lly good form to present on an analytical
diagram any surplus pieces beyond those contained in the
usual outfit, though no restrictions are made, as in the
conventional direct mate, against Obtrusive Men (see
p. 39) within this limit. No. 24 furnishes an example
wit~ one extra officer on either side (see also No. 76).
White has lost only Sand KB, Black only his two S's and
Q. White's KR P, KP and QB P each made a capture
so the QR P, which promoted to Rook, made no capture'.
Hence White's Q SP moved to b3 to let Black QR P in
to promote, and again he moved to b4 to let the new Black
B out via b3. ;-, This new Bishop must have gone around
to h7 before Bla~k's KB P got to g6, and White's KP must
a.lready have b~e~ at £3 to let out KB, so the whole King's
side was lockeq up before Black K and Q got free. Now
Black K bas made at least six moves and Q must have made
three to fall a\ captive on d5. White has only nine p
moves on Quee~'s side, _so to avoid retrograde stalemate
Black must have moved ,last, I Kb3-a2. White before
that moved r. . ~ PC4 X d ; and Black before that 2 .. ,
Qf7-d5. Then Black'~ and Q retract their way home
WHITE. while White P's c/:>me back to b4, cz and dz. A second
pos~ible position '1'.Jrior to · the last nine moves has the
Demonstrate the position prior to the last player's Wh~te P's on b3, 'l':3 and dz, which can be explained by
last two moves. having let out the promoted Black Bishop from hr via
cz instead of b3.

52
... 53
',;,.',
.I
 _j
No. 25.
Each player's last· nine moves may be accurateiy
No. 25. deduced. The ,-'retrogression is : I Qh3 X Q +, Pc5-c6 ;
2 Bc3-e5, Pc4-c5; 3 BeI-c3, Pc3-c4; 4 Pd2xB=B,
T. R .. DAWSON. Pc2-c3 ; 5 Pd3-d2, Bd2--eI ; 6 Pd4-d3, BcI-d2 ;
7 Pd5-d4, Pd2 x S ; 8 Pd6-d5, Qh2-g2 ; 9 Sg2-e3 +,
Original. KhI-gr.
There are two very near retrograde tries which attempt
BLACK. / '3 to get a White S to h7. One breaks into the above play
by 3 Bd2-c3, Pc3-c4; 4 Bcr-d2, Pc2-c3 ; 5 Pd2 x
S=B. This actually releases the Sh7 but leaves White
in retrograde stalemate after Black's tenth last move.
The other goes : 3 Ber-c3, Pc3-c4; · 4 Pd2 X S=B,
Sc2-er ; 5 Pd3-d2, Sb4-c2 ; 6 Pd4-d3 and White S
is successively driven off d3, d4 and d5 by the Black Q P,
for in each case if White S tries to get over to h7, the.
Black; P is obstructed. Altogether, I87 retrograde stale-
mates present themselves in the tries of this position.
Nos. 25A-B, in a somewhat lighter vein, close the series of
examples in _wltich analysis may be said to have no eye
to the future.
I ,f/ V
:/('?',
1
v-0 ..,o. 25A.
,·. , T. R. Dawson.
t, (~----•; Neub~rger Wochenschach,
/'· '7> ...--· -9 Dec., rgrr ..
_..:;:!, ~ . ..._ BLACK.

,/I "l
/
-L- ;,.._;
. WHITE.

Demonstrate the last eighteen moves played to

this position.

54

_-~f¾i.1''~Iq~~ttt@Nt''''·" ___________ _
-. - -
 .,

No. 26.

T. R. DAWSON.
No. 26.
Reading Observer, 26 Dec., 1914.
Really little ingenuity has been shown in the discovery
BI,ACK. of new-motives to justify retrograde analysis in its relation·
to direct chess. The retrograde legality of en passant keys
dates, as we shall presently see, from 1858. The retro-
grade illegality of Castling was demonstrated a year later . .!
by Sam Loyd. F. C. Collins, in the early 'eighties, was
probably the first to compose an analytical retractor. j
These few lines sums up the entire history of retrograde
analysis to date. Later composers have only elaborated
these pioneer achievements. And yet other _possibilities
undoubtedly exist. They always do in chess. It· only
remains to find them. No. 26 may give the clue to a new
class, which we will call the Retrograde Illegality of Con-
structive Tries. This formidable name, like most things
formidable, has a very simple meaning. The solver is
asked to make some constructive change in a position so
that a certain end, usually the mating of Black, can be
attained. A choice of. apparent solutions is presented,
retrograde analysis stepping in to show that only one of
these is legal. In No. 26 we are required to add a White R
in such a way that mate in one may follow. Apparently
the R may be added on abc5, or a6, or d4, each of which
WHITE. allows mate. But the first two of these tries produce an
immediate retrograde stalemate. Black could not have
Where must the White Rook be added so that \Vhite moved last. Only d4 remains, and clearly the position
is legalised at once by a simple retrogression : I Kd5-e6,
may mate in one.
Re4-:d4 +. Q.E.D. !

57
 No. 27 . . No. 27. !
j:
'.
Retract : Pcz x B.
T. B. ROWLAND.
I Qf5+, BxQ+; 2 Pg4+.
Chess Fruits, 1884.
It is doubtful if a simpler analytic retractor than
BI,ACK. No. 27A could be devised. White retracts a single direct
Pawn move, all other suppositions being easily disposed
of by counting heads and Pawn captures. Then self· mate
in administered by moving the same Pawn forward again.
From this elementary po.sition to the more complex type
shown· in No. 27B, we pass through many gradations.
No. 27 is chosen from a number of early examples by the
Rowlands. A pretty point here is that White's. last. move
by P from cz must have captured the Black Q B, for all
other Black men must needs have fallen on Black squares.

No. 27A. Ko. 27B.

F. Dusold. J. Jespersen.
Jamaica Gleaner, 31 Dec., 1892.
. BLACK.

WHITE.

White retracts bi5 last_move and self-mates in two.

WHITE.
·white retracts his last move and ·white retracts his last move and
j. self-mates in one. mates in one.
Retract : 1 Pd3-d4. Retract : l Pd2 X R.
. ! . Play: 1 Pd4. . Play : I Qb2 .
~ '" . \
\)\A.'V,...'-'
• "
v,,vv":\
•.
{--, H -.
...:,; -c>,.= ru
I,.,
~.-»,f I! 1::
' ) .! (. !~. \
• 't 1
59 J
58
 No. 28.
Bolton Football Field, 31 Dec., 1910.
Retract : r Kc7-d8, Pc5 XS + .
Mate by: r QxS+.

The analytical retractor is not to be confused with the

conventional retractor in any of its many forms; Loyd
used to claim that his three retractors in Fiske's story,
Chess L1fe in a Harem (Loyd Book, p. 29), exhausted the
only three themes that could advantageously be presented
in retractor form. An amusing story is told of B. M.
Neill's showing Loyd a retractor of his own, since become
famous· as probably the most difficult unanalytical retrac-
tor, which Loyd proved quite unable to master. The
moral is that the unanalytical -retractor has many more
than three possibilities. Indeed its attractions have been
so fascinating to composers that it. may be the kind has
been overdone. Certainly recent developments of the
retractor, the Twentieth Century brand and others, have
been carried so far as to produce their own reaction. The
artistic skill of Mrs. Baird and the vagaries of some of her
disciples produced some beautiful and some curious results,
but solvers, with the exception of a handful of enthusiasts
in England,. never took kindly to these puzzles. P~rhaps
they left too much to the imagination. Solvers like to
use their wits rather than their fancy. In the analytical
retra.ctor there is assuredly plenty to occupy the wits.
Whether that in itself is sufficient to lure the solver in
general remains yet to be determined.

i
'
!
i
61
• f

'
--- ·-.
 No. 29.·

,...
Retract :1Castling K R. [!. fil_.R1-f 1:i -eA-c.)
Mate by Castling Q R.
s. STEINER.
En passant a~d Castling recur in the retractor as in
.Magyar Sakkujsag,, ~ug., 19_12.
every branch of the Analytic (see p. 37). The uncastling
to prevent illegalities of Group 9 is developed in the present
group into uncastling to permit a new condition, mate or
self-mate. No, 29 is most ingenious, combining retro-
grade uncastling with future castling. The analysis is,
however, more ieco1:dite in No. 29B, though in the latter
the move sequence is inaccurate. In other words, after
two retractions by Black, and only one in between by
Whlte, White is again called upon to move directly. Either
Black must have moved twice consecutively, or else White
is in the process of doing so now. '

No. 29A. No. 29B.

G. Hume. T. B. Rowland.
Jainaica Gleaner, 28 Dec., 1894. Chess Fruits, ·1884,
r,=~~~BL;,;;;,A=CK=·,;;: === ~ BLACK.

, WHITE. WHITE.
It •is ·white's move. Retract It is White's move. Retract
-,:,'/ thE! last ·mov.e on each side and '."\ the three last moves, and
/ White s;lf-mates in two. .' · White mates in two.
. , . Retract : I Castles, Bd5-a8 ; ·
· /I Retract: 1 C a ~Pe7xQ=Sstles, .. 1
2 Pb7-b6. .J
L ·nay: 1Sd6+. , _Play: rRxR+.
62
··.e~~. . . 63
 \

•
6,
'
I
!
'I
.
l
'
I-

r
 -
·--- -·- -- -
~- --- ---- -- - . "-..... -- -~
-- -...---....,~•----...: ::'"'-.::._ --
. ---- -- ._·
-·-

.I
,
<t
.::::_,
,_ I

'O
0
 No. 30.

T. B. ROWLAND ..

Chess Frttits, 1884.

: Retract: Pg2-gr=S.
I f '
Mate by : Q X p. J c..' ,v~,"f i,-11.,;u.,OJ,,v./'j
\ ·. '\-C~/\-V" '> r,
Here again, as in No. 29B, the. time sequence is ~g;;__
I ( .'
~ Black retracts and forthwith White mates. his
means that White, in the game, is moving twice running.
It cannot be that Black has moved twice running, as is
assumed in some of the Rowland analytics, and that the
retraction is for the purpose of restoring the legal alternate
sequence. For in No. 30, after Black has retracted Pg2=S
Black is brought into retrograde stalemate. The P at a5
could not have moved, since the Q R was needed to furnish
the White P's with one of their seven victims. White has
lost three men only, all fallen to the adventurous Pat b4.
Evidently that P could not have moved, or White would
have been in imaginary check.. The P at f6 moved away .
from f7 long ago to admit the B now at e8. The lack of
White men precludes its having come by a capture from
g7, and likewise prevents Black's K S P from having
moved off its own file. We give the position as an an~:y:tic ( r-.._ ,
sound so far as the requirements of the actual condition / {
WHITE. n go, but unsound or imaginary if the analysis is continued : / ~
~Jt""·°" r--furth~ It is at first a little puzzling for us to ~ee just , ] 'I-
Retract Black's last move. Then White mates in one. ' what 1s legal or illegal in retrograde sequence, and un- , '-:,.
fortunately our space is too limited for full explanations , ~-\
of each diagram. The difficulty is like that of trying to i " '-....J
write with the left hand, or to say the multiplication
table while standing on one's head.
i
 - No. 31.
Retract : Rg8 X P.
No. 31. - ~
Mate by P+P_ e.p. + +.
This importaht pioneer of Collins, spoken of on p. 57,
F. C. COLLINS. introduces us to the en passant group of retractors. No.
31 has an e.p. key made possible by White's retraction,
Frontispiece, Chess Problems, 1881. and made legal by Black's only proper previous move, Pg7-
g5. That White's retraction should be analytically sound
BI,ACK. must have been almost beyond the comprehension of the
-early solvers, and it trips many a good analyst to this day,
who· sees in the curious grouping of the White pieces only
a bizarre pose. The apparent freedom of the White force
is illusory, as White's last move must have been a capture,
and all captures save that of the KS P have been made
by the P's now safely blocked in on the QR file. Nos.
31A-B have an en passant capture in the retrogressian as
like Nos. 19-:-20. No. 31A is not purely analytical, as the
last two retractions are partly determined by the need of
obta,ining a Black mate in one, thus : 1 Kf6 x P, Pf5 x
g5 e.p. + ; 2 Pg7-g5, Sg5 X Q ; 3 Qd5-f7 +. Black
could have mated by Qg8 + +.
No. 31A. .No. 3IB. A. Troitzky.
V. Onitiu . V., Magyar Sakkujsag, Mch.,1913
.lvlagyar Sakkujsag, June, 1913.
BI,ACK.

WHITE.

White retracts his last move and mates in one.

It is White's move. Retract

the last two moves on each side
WHITE. and White mates in one.
It is White's turn. Retract the Retract : I Kh4-g4, Ph.5 X
last :live moves and Black P e.p. ; 2 Pg7-g5, Rf6-h6,t·
mates in one. Play: I Rf4.
66
 .,

No. 32.

S. LOYD.
Musical World, 1859 ? No. 32.
BLACK.
I QaI.

Two main divisions of retrograde analysis remain for

our consideration, the en passant keys and the Cant
Castlers. The former are considerably the more numerous
and in some respects the more complex, so we shall take
up the Cant Castlers first. They originated, only a year
later than the en passant keys, in Loyd's pioneer, which
to this day remains, like so many of Loyd's pioneers, the
most economical example. The theme is in its simplest
form : Black is in retrograde stalemate, except as to his
King and Rook. Consequently one of these two pieces
has just moved, and Castling, though it would defeat
White's key, remains an illegal_ resource. The greater
part of the early examples· showed this same motive, just
as the en passant key ran in one conventional groove
practically from r858 to r908. As early as r886, however,
a new idea was infused into the analysis. The late F. C.
Collins, the composer of Nos. I9 and 3I, again came to
the rescue of originality with No. 47. No other new argu-·
ment was invented until r905, when H. W. Barry published
WHITE.
No. 38. Since then several composers have produced
fresh themes, N. Hoeg, J. C. J. Wainwright, W. Pauly,
Mate in two. M. Marble, A. Bolus, C. S. Kipping, A. J. Fink, and others,
(. culminating in the productive and co-ordinative outburst
of T. R. Dawson, discussed on p. 73.

68
6g

. ll
{- I·.
 No. 33. No. 33.
I Qf6.
P. H. \VILLIAMS.

Chess Amateur, Aug., 1910.

The Black P at h3 has made one capt{i.re and the P
a.t f3 has made three. As White has lost four men neither
BLACK.
the P at 8,4 nor that at b6 can have moved last, so that
Black has clearly lost his privilege to Castle, by moving
King or Rook last. This example, while not very intricate,
caused such a commotion among the large body of solvers·
in the Chess Amateur as to arouse the latent interest in
analytics, causing a. widespread increase in such com"
position in England. In passing let us notice No. 33A,
by a leading composer of the last generation, and No. 33B,
with the prettiest actual play of any Cant Castler.
. .

Ko. 33A. No. 33B.

H. Lehner. D. W. Gilmour.
Oesterreichische Schachzeitung, • Glasgow Herald, 1 June,
Dec., 1.875. 1901. .

BLACK.

WHI'.J'E •

. l\fate in two.

WHl'.l'E.
j
Mate in two. Mate in three.
I fXQ, 1 Bg~, Rh4 ; 2 Be4.
 No. ·34_- ,
..
\

T. R. DAWSON.
.No. 34.
Ill. London Ne'/£/s,

BLACK.
20 Dec., r9r3.
r4tt-. 1.R f1 i ·,
i
.,
l
;l

The year 1914 saw, in America, the birth and death '

~V1--
■ ~14 ■
~•~
of a unique experiment, the establishing of a weekly
magazine devoted exclusively to problems. It was appro-·

■
priately called The Problem. The experiment was doomed
14
}/////////✓/~f{~f&,,,//2~-- ~-- I
to failure, in spite of the earnest efforts of the editors,
H. L. Dolde and C. P. Carpenter. They certainly _tried
~~i• ■
~'//2~~~

ia their best. The most serious feature of this little paper

was the publication, in instalments, of T. R. Dawson's
analysis of Illegal Castling, which changed the standing
of this type of retrograde analytic from that of a sporadic
eccentricity to that of a recognised and classified study.
Dawson has composed about one hundred Cant Castlers,
and this section of the present book is based altogether
on his article and largely on his own problems. These
range from the great simplicity of No. 34 to the complexi-
ties of later groups. The solver often enjoys by preference
tbe simpler examples, and positions like Nrn;. 40, 46 and
48 will make converts where the more recondite specimens

• •- iw,,j,,,~
WHITE.
■1 arouse only apprehension. Certainly the White pieces
in No. 34 cannot be accused of congestion, yet nothing
more is necessary to show that White King or Rook has
moved. An extension of the retrograde ,stalemate theme
Mate in two. occurs where the stalemate can be demonstrated to have
taken place in a position prior to that of the diagram, this
itself then showing apparently that the castling side is
quite free. ·
j-

72 73.
 , C
re
.;,
vi~·-;
1~.i'-t ("• 'l) -
f 0t . 9-,, \ ' ,;:-'- J,,rM . -
I
i . Q, -v
v'V ,f 'D
V, ~ .#. ✓ v-r·
,.,/,I
. • u ' / '/.• .·, I
.
·,, ·;
I , , ,,ti ij1;,
I ".,.
, , ,/J . I11,; I: .
.-·
No. 35. No. 35. /
·; . ·"
• ,\··
-. i .
l . l l
.., / .- t 'I,:
There are several practicable methods of showing that .
T. R. DAWSON.
the retrograde stalemate occurred at some stage earlier
than the diagrammed problem. Thus in No. 35, the Black
Natal 1lferrnry, 22 Nov.1913. Q R and one S were captured at bJ and c3, and the K B
and other S crowded in the corner. All this must have
oc-curred before the White Q B was free, and he then had
to travel via dz, er, g3, d6, c5 to b6, where Pc7 X B, releasing
the Black- Q. But during these six White Q B moves,
Black had only five P moves, Ph7-h6-h5 and Pg6-g5-g4
-g3. At the crucial moment when the White B reached c5,
these moves were exhausted and Black must have moved
K or KR. Nos 35A-B similarly illustrate the release of
Black's QB and QR one move too late to save the pre-
sumption of Black being still at liberty to Castle, so White
can mate by I Qe5, or I Qe6 respectively.

No. 35A. No. 35B.·

T. R. Dawson. T. R. Dawson."
1913./ / Magyar Sakkvilag Ty., 1914.
0
====
BLACK. 14

--'-..;....-:;;.;......,;;::::::;;;i~:-------------.......;,.,___..-.,..t..,c..,._t,.__ . .~·--·---- .
-~---
 .. ,...• ~. -·•"'-•'• ,._,,

'

No. 36.
.\

. T. R. DAWSON.
,J I/ ;J ' No. 36.
I"
The _Problem, 4 July, 1914.
The retrograde block is particularly well hidden in
No. 35B, in which several of Black's pieces appear to have
sufficient freedom to justify Castling. But until Black
moved Pb7 X Ra6 the Black Q and Q B were completely
housed on the home squares. The Q R went to b8 and
the S went to a8. The _other Black S sacrificed at c3, and
the White B from cr went to b6 .. Meantime White's
Q RP got to a6 and K R to ar. Then Pa7 X B and White
played Pa6-a7, and Black has blocked in the QR where
it has no move. White plays Rar-a6 just one move too
late. These Castling Analytics, which the editor of The
P1'oblem referred to as Post-Mortems or Inquest Problems,
also remind us of stout gentlemen arriving on the railway
platform as the train is drawing out of the depot. They
are always one move too late.
No. 36 shows how a free Black piece, instead of having
been blocked, must have been pinned by White at ·some
earlier period., The demonstration is not an easy one to
present in orthodox form. The K B and consequently
the White Q cannot have reached g8 and h7 until White
had played Pez x Sd3, closing up that sardine can in the
south-west corner of the arena. The P at £5 moved out
long ago to release Black's K R, now among the sardines.
Therefore the K_B and Q had to enter via g6, pinning the
Mate in three.
f_- Black B at f7 as
they came or else forcing the Black K-
! to move out of check. Either of these explanations robs
him of the prerogative of Castling, and White mates
prettily by playing Qh8.

77
 T. R.
No. 37.
DAWSON. ~~
f No. 37.
\ , Retrograde stalemate may also be caused by having an
Gazette Times, 2I Dec., r9r3. , apparently free Black piece blocked by a White piece or
~ Pawn. In No. 37 White has lost KR and one S, captured
~ by Pf6 and Pc2. Hence Pd4 had to go up long since to
ti:.
\
t. ~
let KR out, after which Black's KB reached c3 via b4;
which means that Ra3 had to be on a2, Rb3 on a3, and
~ Pb4 on b3, so that after White moved Pb3-b4 Black's
1 only free men were. K and K R, one of which had to move .
.: ;) In No. 37A White's KB P promoted to Q on g8, and in so
.~ i. doing blocked the KR. In No.. 37B the White S at c8
\ 1 entered via e7, blocking B~ack's KB and forcing the KR
3 ""- to move.
1'.:i~
/ ~ No. 37A. Xo. 37B.
? ~~ .• T. R. Dawson. T . R . D awson.
~ . '1 V., Bolton Foo_tball Field, . The Problem, IS July, r9r4.
, .._ , ~ 20 Dec,, r913. 3
{ .., 1
l X"'1 ' BLACK. ==;~ . BLACK. . .J. ...
• ! ' \.)
,,
.' '\_)) <,-
\; ~
l -,.I";>.
·.L'\~
t ,"-.
WHI'tE. If. '~
~ ,l
;
~ ~
Mate in two'. I
I
1, '-,<::l
v
I ~ "'
if)'§ .
. .. "'
 No. 38.

H. \V. BARRY.

Brooklyn Eagle, 22 Oct., r905.

No. 38 ..

BLACK.

The Castling try is illegal, because the White K has

moved to· admit the QR to e2. This clever problem was
the first to demonstrate a King move as the preventer of
Castling, and it played great havoc in a solving tourney,
as so many of the pioneer. analytics have done. .There
is quite a romance to be written about the origin of each
of the analytic themes, and it might remind us of the
romance of a summer's dawn, which contains in itself the.
element that is about to dispel the feeling of mystery and
novelty it causes. There is no longer any romance about
the analytic. The tourney solver is no longer caught by
its strangeness or its unexpectedness. He may be con-
fused by its subtleties or complexities ; but the light of
day has rendered familiar the nature of the phenomenon,
and the length of time necessary to master any particular
example will depend solely upon the _solv:er's. analytic
skill, no longer upon a happy chance msp1rat10n. The
passing of romance in the problem world opens the road
for classification and the scientific docketing of themes.
'WHITE. The change is as inevitable as are all changes depending
on growth, and its advantages or disadvantages are largely
Mate in two •. a matter of taste ; but we can well afford to remember
the romance of the past, and the many positions which,·
like No. 38, bear testimony to its having been.

80
I r~il,~ ti :=i- (1<- , ,11
-~: ..---." ·"'
·'
.-.,-,--:.'-',"•

i f[.J,,
,,.,_.
1-w,v,
·-· _;....·"
_ /
1
No. 39. ··--- No. 39.
I Kdz,PxP; z' Sg7.
J. C. J. WAINWRIGHT.
Here both Kings have moved, and neither side can
Gazeite Times, 3 March, 1912. Castle. As in Nos. 39A-B, the Black K has moved to let
out the QR. In Ko. 39B, the QR cannot be explained
BLACK. as a promoted P, because only c;me Black P is missing and
it has been promoted to QB. The White K in No. 39
was mov"ed when Black's KP promoted to QB at dI
checking at dz in transit. In No. 39A the K moved t~
allow KR access to cz, or perhaps it was the QR who
moved. Either explanation nullifies the claim of a right
to Castl~. ~hese t~o positions lead one to expect the
No Castling sign agam, as to No. 39B, but \Vainwright has
cunningly completed his trio by making Castling legal
and very necessary for White.

Xo. 39A. :N'o. 39B.

J.C:. J. Wainwright. J.
C:. J. Wainwright.
Reading Observer, Dec., 1912. Brooklyn Eagle, Dec., 19u.
BLACK. BLACK.
===

WHITE.
Mate in three. Mate in three.
1 Pd3. I Castles, Rg7 ; 2 P x· R.

/
-
 No. 40.

J. C. J. "\VAINWRIGHT.

Am. Chess Bulletin, April, 1912. No. 40.

BI,ACK. I Bb3, KbI ; 2 KdI.
There is a good try by I Pb4+, but White cannot ,.
Castle as the K must previously have moved to allow his
rival admission behind the lines, via dz or eI. The position
is one of those graceful and economical ones which conceal
the analytic intention and win the solver's appreciation
without appearing to strive for it. The theme, however,
has not the depth nor the difficulty of the more complex
speciinens. The attitude of the solver in respect to these
more complex analytics is not always a patient one, and
reminds us of Samuel Johnson, One day, we are told,
Dr. Johnson was in a company listening to the efforts of
a great· pianist. At the close of the piece his neighbour
turned to him and said : " That was certainly a most
difficult piece to play." "Difficult! " answered Dr.
Johnson with an exclamation, "Sir, I wish it had been
impossible." Now the reluctant solver of analytics who
accepts the more complex retrograde studies in no better
spirit than this really is condemning himself. The lack
of patience is his and not the composer's. The field is
WHITE. comparatively a new one, offering totally new points of
view and totally new values. If at the first plunge any
Mate in three. sense of confusion ensues it may be the fault of the reader
as well as that of this book, and there can be no harm in
making another try to get a full mastery of its contents.
' .

85
 •. I
I

No. 41. No. 41.

A. J. FINK ..
In Groups 38-~ the Castling King had forfeited his
Gazette Times, 28 Dec., 1913. rights by moving to allow passage for one of his own men.
In Groups 40-42 he moves to oblige the enemy. The
BI,ACK. pretty Castling try in No. 4r is illegal because the K moved
to let Black's KB reach cr. In No. 4rA he moved to let
the KP promote at er. In No. 4m the KB P promoted
at fr, checking at fa in transit, recalling the Cant Castle
reason for White in No. 39. Clearing for the adverse P
has indeed been one of the most popular themes in the ·
Cant Castle revival. Sometimes, as in No. 4JJ3.,.liis~
that the K or R has moved for the convenience oI a P,
without its being determinable, or important, just which
of the two it has been.

~O. 4IA. No. 4IB.

_-\_ Bolus. C. S. Kipping.
V., Birmingham Post,
25 June, 1912. Re~ding Observer, 21 Dec., 1912.
BLACK. BLACK.

WHITE.

Mate in two.

WHITE.
?.fate in two. Mate in two.
1 Qez. l RxP+.

86

.......
 I
No. 42. ··'

T. R. DAWSON. No. 42.

1 Kfz+.
This position is a modifica9on of the last two groups \
in that the Castling King can be shown to have given .
gro~n1 before an adverse promoted man.f The only pie~e / '
Whites Q B P can have" taken was a S. Hence Black s ~ )•
Q B P becam~ a S after! the (tangle) on th~ left s~de was00,tr ~ :) I
reached, and 1t must have checked at d3 m gettmg out. Id · ') n
One of the two Black S's in the diagram.is therefore obtru-
sive. Such a difficult theme to illustrate does not lend
. itself to a combination with any great merit of solution.
It is sufficiently remarkable that any non-castling direct
solutions can be worked in to all these varied analytic . j'
positions. The direct solution and the Castling try are i
not absolutely necessary adjuncts of every Cant Castle
analytic. In a refractory case it is enough to write :
" May White, or Black, Castle ? " and we shall be back
in the domian of pure analytics (Nos. 19-25). · Fortunately
most themes can be given what we will call orthodox form,
and, as we have a much larger number of positions to select
from than diagrams, we need not resort to the disembodied
analytics at all here, and only seldom in the en passant
key section (seep. 153). In the same way we have been
WRI'.l'E. able to omit all positions· supposed to have arisen from
games at odds. But few of our examples will have the
Mate _in two. merit of solu.tion or of try that No. 41 had. In some cases,
indeed, as in No. 34, solution and try will differ only in
{; t
~ I . t ~
the small· particular of whether the K may or may not
- -{\¼~ ~()_/r I!\/,~ l j jump over the R who moves· up to his side.

Vt<it51 .
I
88'
~r-r e~ T / ~ ;:·
,''
·"
r j.
 I
\
[ t·.
I

No. 43.
No. 43.
I Rdr. i.
T. R.· DAWSON.
Another sub-section, admitting of very interesting
· Gazette Times, 21 Dec., 1913. motives, is that where the Castling King had to move to
release a retrograde stalemate. Thus in No. 43 we can
show that the K has cleared for the Q, the last moves
having been I Pc7-c6, Kdr-er ; z Rg3-g2, Qer-fr +.
This is not the same thing as the clearances in. Groups
38-39. There the K had always to move from positional
causes. Here, however, his moving is necessitated by
time pressure ; if Black had unlimited time there would
be no need for the K to move at all. Examples have been
composed showing clearance by th K16i. each of his own_.
1
pieces,·and for ea:clropposing piece.~ No. 43A is an instance
of· releasing··an-opposin:g-·p:·- ··
.
11,c ., ... {...-
~
No. 43A. No. 43B.
T. R. Dawson. T. R. Dawson.
V., The Problem, 29 Aug., 1914.

I
!

Mate in two.

WHITE.
1
i- ~fate in two.
I Kfa.

90 91
 No. 44.

T. R. DAWSON.

Hampstead Express, 27 Dec., r<;13. No. 44.

BLACK.
I RdI.
Returning to No. 43A, we soon see Black's last move
was Pb6-b5. Before that the P's a7, d5 captured all
Black's missing men, so Pb6 had been in position some time f
.J
to release the Black QB now, apparently, at dr. As a
matter of fact, however, the QB must have been sacrificed
to a White P and the BdI is Black's KP promoted after
one capture-of White Q. Consequently. the last moves
were : I Pb6-b5, Kfr (2)-e1 ; 2 Pdz . B, Kgr-fI (2 +) ;
3 KdI-cI, Bb1-c2 + ; 4 KeI-dI, S-f3 +.
Another way to relieve retrograde stalemate is by
uncapturing, as in No. 43B. Here the last moves have
been : I Ph7-h6, KfI X B !
Finally, the K, under time pressure, may be driven to
cover- a line of attack on his rival. Thus, in No. 44, White
cannot Castle because Black has just played Rb6, preceded
by Kfz--,--eI discovering check.
Looking back over Groups 38-44 we realise how varied
the demonstration of the Cant Castle theme has become
. in each of its main divisions. In turning now to the past
\"\"HITE. history of the R we shall find repeated all the sub-divisions
of the K groups, wi~h one additional idea, the transposition •• I

:\late in two. of the two R's. But first let us stop for a moment's relaxa-
tion with the amusing situations of Group 45.

92 93
 No. 45.
No. 45. .
\Ve are now so e.xpert in recognising how the K enters
T. R. DAWSON. into Cant Castle analysis that we. see at once, in No. 45,
I
that he _has moved to let out the Q R, :whence r Rgr ; ; I
Tagliche Rundschau, 24 Dec., 1913. 2 PX P 1:' th~ 01;1-ly solution. By changing the P from d3 . i
·, l
to d4 this solution does· not effect mate, and as Castling
BLACK. i~ now leg'.1-lise~ w~ resort to)t ~nstead. A pretty adapta- !
\ tion of this twin 1de8: to Blacks force occurs in No. 45A.
H~re Black w~mld be 111 retrograde stalemate were Castling
still legal, which cannot be. So we play I Ra8 and ·mate
. : i
\., .
''
I·

t next ~ove. If w~ add a Black Pat g2, however, Castling

becomes legal aga111, so that now I Be5 is the only solution.
I· We_ ~ive ?ur la_st dia~ram, ~o. 45B, to another amusing
pos1t10n, 111 which neither side can Castle, owing to the
necessary clearances for Pawn Promotions, in spite of the
very great opportunities offered.

• No. --l-5A. Xo. 45B.

T. R. Dawson.
W. Pauly.
V., Bolton Football Field,
1913. 7 Sept., 1912.
BLACK.

WHITE.

(I) Mate in three.

(2) Put Pd3 on d4, and again mate in three.

(Ji) Mate in two. :\fate in three.

(!)' Add .Black Pg2, and again
mate in two. . 1 Kfl, Rg8; 2 Rei+.

94 95
 No. 46.

T. R. DAWSON.

Nol'thern Whig, 6 June, r9r2.

No. 46.

BLACK. f r Rgr, Kg3 ; 2 Rhz-hr.

After our little digression in the last group, we return
'to the more accurate classification and begin on the R
classes. But although we have said that the R can furnish
all the same Cant Castle motives as the K, and one more
too, we find that this section is not nearly so extensive as
that dealing with the K. This fact is due to its being
extremely difficult, if not altogether impossible, to make Q,
B or S react to any great degree upon the corner squares
· of the board.
No. 46 corresponds closely to No. 38. The R on the
second rank has had to get in behind the P's, dislodging
the Castling piece on the way: Or the position may be
explained on the theory that the two R's have changed
titles, the original Q R being now at hr. We consider
this explanation less plausible, although it involves less
moves, because there is an assumption in these Cant Castle
problems that both King and Rook will make every effort
possible to retain their home squares, only moving under
unavoidable compulsion. Consequently if we find a R on
hr we feel sure it must be the K R unless there is a definite
WHITE • .
.. demonstration to the contrary. We shall see in Nos.
49A-B that cases do arise where we can be convinced that
Mate in three. a R has been dislodged and replaced by his colleague from
the other corner of the board or even by a promoted Pawn.

'

96 97 G
 . i
M
.,Ii ·11
No; 47. f No. 47.
.' ••f i:

F. C. COLLINS. r SxB. ' f ;

I
· · 3 Pr., Sheffield Independent, 12 Mar., 1886. Loyd's No. 32 had two or three early imitators besides I
BLACK.
the fine elaboration by Lehner already quoted. There was
even one by Zukertort. But it is much easier to imitate
l
than to originate, and until F. C. Collins published No. 47
nobody thought of :finding any new reason for the preven-
tion of Castling. Even to-day we do not see at a glance
in his problem that "\Vhite's QR P must have promoted
on a8, so ~hat Black cannot C:1stle on his Q side, although
~e may still ~o so on the ~ side. The problem competed
m a compos1ng and solv1ng tourney, and the havoc it
played with the solvers doubtless helped it to a prize it
would probably not have won otherwise (see p. rr7).

No. 47A. Xo. 47B.

N. Hoeg. W. Pauly.
Shakbladet, Dec., 1906. l9II.

. , I // WHITI\ •
.,·. : 'Vl.,•_0· l-,.,d "ti,.;., Llr ~
· Mate in two.

WHITE.
Mate in two. Mate in two.
1 Rd1.

98 99
 No. 48, 1
lf
l R. DAWSON.

V., The Problem, 3r Oct., 1914.

No. 48.
In this problem the Cant Castle analysis only prevents
a dual, not a cook. After I BXB, KgI, White must
continue 2 Kez mate ; for in admitting the Black B to
its odd position he has forfeited his right to Castle. But
the economical rendering of a recondite theme more than
compensates for running so ·near the wind of orthodoxy,•
as we defined it on p. 89. The retrograde moves leading
up to the diagram have been about as follows : I RbI-aI,
Bar-bz; 2 Rb2-c2, Kgr-hr; 3 Rb4-b2, Bbz-ar;
4 Pc2-c3, Bc3-b2. The position is an excellent example
of the R moving to make way for an opposing man, corres-
ponding to No. 40 in the K series.
Group 47 also corresponded exactly to Group 4r, and
the popularity which we saw at~ached to the K making
way for the promotion of Pawns is equally present here.
Indeed there is very little difference between the two, and
in such a position as No. 47B is it quite impossible to decide
whether it is the Kor the R who has moved. As in No.
47A, it is clear that one of the Black R's is obtrusive, for
the R at h8 never got far from home, but in No. 47B the
promotion may have occurred on any one of the eight
WHITE.
squares on the first rank. Whichever square it was on,
.. K or R moved either to give ingress to the Black P or
egress to the new R. The problem enjoys no less than
Mate in two. five variations, quite a number for an analytic. It would
l
·. !
indeed be a gem were not the Castling try so closely akin
to the actual solution.

IOO
IOI
 No~ 49.
I Rfr+. .
R. DAWSON.
Re,e the R fr .· ·
Magy ar Sakkvilag T y., 1 914• attack of o£ the
manreuvres ompieces
hr was
f re qmced
. to be at , .
ready t the Bh6 on the Borl position to co g7 during the
. come to ack K , ve, the li
ite R
0
and r l g7. This Shad until the Black Sne of .
no example f e ":"" the Be8
Wh to be on dB t was
specimen at~ t1ris line covei In the K secti o cove, the
,by Nos
Dawson . and,. butthe the:::"'has we hrul
e as since b n~n o,thodox
we gave
the R~ot9A-B c_orrespond t O G een developed
4
mot'10n as moving
b f t o releas roup 3 m • that th
not fro~ pos1t~~!'aI'~ing requ1re:tl~o~adt~imestalem!tesh~r
auses. •
l)ressur e' ands

No. 49A
T .
V . R. Dawson
., The Proble m, 28 KoY
. I
BLACK. ., 914.

WHITE •
. j- Mate in two/'-
Sh+ . . 1
i) .
1
/ ; ' Ke,/ ,u};,~
I02 I" c.vJc,_..:, ,_,. 0., t1M{J:' ·J,,Jci;; I ~

-
 · .. ... s >' t;·

No. 50.
T. R; DAWSON.
No. 50.
The Problem, 28 Nov., 1914.
r QxS+.

· , Perhaps this position, taken in comparison with No .

. 49, sliows-as-well as any the difference between till!e _
PJ:..§.§.UX,~ and a p()sitional.£ause,_ Here the last moves
have· 15een something like this: r Ph7-h6, Ra5-ar;
2 Sh5-g3, Re5-a5 + . The only way we know that such
a series must have taken place is that any other would
promptly leave Black in retrograde stalemate. If Black
only had other moves, a free B, as in No. 49, for instance,
then the position might have been arrived at in a dozen ·
other ways, without moving the Q R at all, and Castling
would be quite legal. In No. 49, on the other hand, there
was no time pressure, thanks to the several free Black
men, but the R was forced just the same to mask the
diagonal line of attack on the Black K, simply because no
other piece could do it.
The release of retrograde stalemate in Nos. 49A-B is
not difficult to trace. In the former the last moves have
been: r Pf7-f6, Rgr-hr ; 2 Kh3-g3, Bhr-gz +, and
everything is comfortable again. This parallels the strategy
of No. 43. In No. 49B we have the same theme of un~
capturing which marked No. 43B, for the last moves must
have been : r Ph7-h6, Rfr X B !
:Mate in two. ~ • So far we have examined the R motives which parallel
vk0-o-wvvd. (.:tJw,v»...
1i,;i-p)4 J w-r~.1~ .- ·
those of the K. There remains·only one more character-
istic division to examine, which depends for its possibility

- ~ ~ "r 1-o "'G, 1 c 2 :.o d4 _7Ze,..w:, v-t R °" 8. I

I
'
on the fact that there are two R's in each set of chessmen.
.

i
1,
I;.
\.\
105
_,. l,.
 I.
t
/
[
No. 51..
I Qa7.
Reference was made on p. 93 to the one Cant Castle
theme which the R can illustrate, but not the K. This
is the transposition of the two R's, clearly inapplicable to
the K, as there is only one Monarch in each army. In
No. SIA the Rh8 is Black's QR, the KR l~aving been t
enclosed at h3 before White played Pfz X B, which released

.
i S
the White Queen, allowing the subsequent release of . iI
B1~QB,. · In No. SIB White played PdzxS, followed ;

C. v~
afte! ~n interval by,,.,Blac.k's Pa7x B, followed again by
Whites Q Rh3-f3;\ ·then by Phz X R, and finally by
K Rdr-ar. N"o-sr shows a promoted R at h8, one of
]?j,:7. ;<.1!(6/Pack's ~•s being at fa and the other having fallen at I
!
:
i
·,.J
. :
1
✓ gJ, allowmg the KR P to promote. "J.vv·~ ,:,lN',...,e,!.,o di_,. tt~i
I ( j . . !
No. SIA. No. SIB. !
T. R. Dawson. T. R. Dawson.
Natal Jiercury, 8 Nov., 1913. JY'atal Mercury, 13 Jun~, 1914.
BLACK. BLACK.

Mate in two.

106

-------
 ..-
No. 52.

P. KLETT. No. 52.

Schachproblenie, 1878.
f
I PxP e.p.
BLACK. There is a great difference between the Castling analytics
and the en passant analytics, to a study of which we
have now arrived. The Cant Castles, if we omit the work
of T. R. Dawson, could almost be counted up on one's
fingers. From its origin in r859 the theme remained
virtually unexplored for fifty years. The en passant
keys, on the contrary, appeared at regular intervals every
year or two. At least one main type, which on p. rr5
we shall define as the conventional side square type, did .
recur constantly. The other classes, first disco.vered by
Loyd in r894, did not come into general study until r908
and afterwards. But the conventional type throve. It
had the backing of great names, as we shall see from the
headings to some of the next twenty diagrams. And
there are many others, here crowded out for lack of space.
· -- Our own collection now includes ro5 specimens, of the
conventional type, by 85 composers, and it -is certainly
far from complete. There is Loyd's Spectrum Analysis
(No. 66 of his Book), and one by A. F. Mackenzie.~ Shink-
man is represented, and Pradignat, and Galitzky, and the
Bettmanns, and J. Moller. Juchli has a complex example
WHITE'. (No. XII. of his Book) ; Kipping has a Block-threat
(White to play, No. 84); and there are two examples in
Mate in two. Tasks and Echoes, No. I5B, by K. Erlin, and No. 24B, by
W. Wolf, with its remarkable clearances. But no name
has given the theme mo.re distinction than that of Philip
Klett, Germany's master problemist, who composed two
widel); known examples.

10S 109
 -. > ,- •

-,:,_":··:--": ~--!f!> -, ..~- ~. -

No. 53.
No.' 53.
.I PxP e.p.+, Kd5; 2 Sfz.
ANONYMOUS.

Schachzeitung, Feb:, 1858.

, The actual origin of the en passant analytic is not
certain. Way back in 1844, a year before the publication
of the Indian Problem, old Portius, the genial editor of
BLACK. the Leipzig Illustrated, published No. 53A in connection
with one of his chess tales. It is simply a trick problem,
quite free of any thought of analysis, but we quote it for
its historical significance. It set solvers' brains in action,
and suggested the possibility of an analytic en passant
key. A good attempt was that of Schmidt, No. 53B,
which unfortunately is an illegal position owing to excessive
\Vhite Pawn captures. No. 53, six years later and fourteen
_years after No. 53A, is analytically sound, but unfortunately
unidentified.

No. 53A. No. 53B.

K. J.
S. Portius. C. Schmidt.
24 Aug., Leipzig Ill. Zeitimg, 10 Jan.,
1852.
BLACK.

WHITE.

Mate in three.

Mate in two. Mate in four.

IPxP e.p.+. IPxPe.p. +.

IIO III

I
I
. I
'-----'...c.,·.=,...1':..-·..::_..___:__.:.,._,.___ _ _.:_____,,---~~----:__--,c....:...·:.......· ~t- ------~_.__.,..__;.....:....,.._________;_._.....,....._..,...,,.,,,,,
 . .
No. 54.
No. 54.
E. B. CooK.
Am. Chess Nuts, 1868. , I PxP e.p., KxS; 2 Qg4.

The foundation of our general classification. of en

BLACK.
passant analytics will be laid on the theme Pawn's terri-
tory, and it will be convenient at this point to define our
nomenclature. Nine squares constitute the thematic
· arena. The " original " square is that one on the seventh
rank from which the theme P, who is about to be captured,
has just moved; the "passage" square is that one on the
sixth rank over which it passed; and the "present"
square.is the one it occupies after its double move. There
are further three pairs of squares divided on either adjacent
file over which the theme Pawn exerts some influence.
The most important, and historically the most important
in the whole arena, is the pair on the fifth rank, which
the theme Pawn attacks when it stands on the passage
square. These are the " side " squares. The remaining
four squares are not of sufficient importance to require
special names.
It is always a simple matter to exclude the theme Pawn
from having just captured on the last move from the'sixth
rank to the present square. Therefore the whole of our
classification divides into two groups : we may prove
that the theme Pawn cannot have been on the p·assage
\VHITE. square or we may prove that it must have been on the
original square. In the first case we have to prove an
Mate in four. absence ; in the second, a presence. Since the presence
can never be proved without taking into account 'White's
1· preceding move, the proof of absence is logically nearer
and we commence with the various groups illustrating it.

·112 IIJ H
 No. 55.
No. 55.
I PxP e.p., KxB; 2 Ph7+.
B. HARLEY.
The simplest Cant Castle idea was for the Castling side
Westminster Gazette, 1902. ' to be in retrograde stalemate excepting the King and
Rook (Nos. 32-34). This necessitated King or Rook
Bl,ACK.
having just moved, so forfeiting the right to Castle. The
simplest en passant idea is for the Black force to be in

■ ■
retrograde stalemate excepting the theme Pawn. This
'11~' .
··•·■-■
t I requires the Pawn ·to have made the last move. Then, by
placing the White K on a side square, the theme P is

----■~,~~
forced back to its original square, since on the passage
w.. square it would at once give an imaginary check. We call

■ ~. Y'"~W,~m,,~I
■ ■ I i 11
this simple arrangement the Conventional type (~os. 52-

......,,,1
62), and Ko. 55 may be studied as a most economical
· rendering.
No. 55B .

•~-·---··
Ko. 55A.

■.■.•.•1
H: A. Loveday. F. Amelung.

I Ill.London News, 30 l\larch, 1872.

Bl,ACK.
Diina Zeitung, 1897.
BLACK •

........
■ ■ ■ ■ WHITE.

Mate in three.

WHITE.·
· Mate ,in two. Mate in two.

. I
1 PxPe.p. +. l PxP e.p .
i
I
. I
. I ,
II5

..,.,,
lr
I
.1
 ,'j

No. 56.

A. MULLER.
, No. 56.
I PxP e.p., PxP; 2 KxP.

The simple settings of the Conventional type have been

very frequent. One of the most interesting historically
is No. SSA. by the author of the Indian Problem ! Un-
fortunately it is analytically unsound, as Black's last
move may have been Kb7-a6. An unconscious revision
by Amelung, No. 55B remains the most economical of all
en passant problems. It will easily be understood that
solvers, critics, and last, but not least, those dominating
factors, tourney promoters and judges, eventually took an
unfriendly view of the many permutations of this conven-
tional theme. The petrification of the theme was perhaps
hastened by tourneys, which distinctly forbade e.p. keys,
and later took their exclusion for .granted. It speaks well
for the virility of chess ideas that the e.p. analytic, in spite
• of the official embargo, has gained new forms of life, so
that now-a-days the tourney prohibition rests "more· on
convention than on any reasoned argument. ·
At its best there is no lack of cleverness even in· the
simplest Conventional type. No. 56 is entirely berNt of
officers, yet it conceals a very interesting retrograde try,
as to all appearances the last moves might have been
I Ke8-f8, Pg6 X f7 +. This would require seven captures
on each side, and each side has lost seven men. Only a
Mate in three. nearer study will show that in this try all of Black's captures
must have been on Black squares, Pa7-fa, Pg5-f4, Pf6-
· g5, and one of White's missing men, the KB, never left
the White diagonals. Hence the last move was Pg7-g5.
 No. 57. No. 57.

M. LANGE . I PxP e.p. mate.

. i
J.
. It is frequently stated that en passant analytics can
f einheiten des Schachspiels, 1865. be grouped according to the simplicity or complexity of
their themes. By this is meant not so much the amount
BLACK.
/3 of material used as the completeness of the analysis
required. No. 57 looks very niuch more complex than
No. 56, yet the latter required a complete explanation of
• 1
the position of each P, while in No. 57 the only point not I
clear is why Black cannot have played K X B. This !.
becomes clear when we notice the P at g3, which has 1
1
captured the B earlier in the play. Consequently the Black
. · K is in retrograde pin (see p. 45) and cannot have made 1
the last move. j
No. 57A. No. 57B.
11. Lange. J. N. Babson.
Handbuch, 1862. B.C.M., Dec., 18~3.

WHITE.

Mate in one.

WHITE. WHITE.
:Mate in one. Mate in two.
I PxP e.p. I PxP e.p.+.

' II8 II9

 -··. ~ ~ -,-. ·--
' ~- ·1, "C:i'"' • ~ ., -.,;. • ·~'7¥:..

No. 58.
A. NORLIN.
No. 58.
V. Tidskrift, 1907.

lll,ACK.
r I PxP e.p.+.

10 The childlike character of the simplest en passant

keys provided little food for thought for the solver.
Although Lange and others tried earnestly to complicate
the analysis, it was not till I876 that Loyd's Spectrum
Analysis required an explanation to be found for the setting
of the entire force on both sides. Now-a-days such com-
plete analyses are still rare in the conventional type, and
we look for them mainly in the newer branches of the
'
j classification. The usual examples of the conventional
j I
type contain single interesting motives rather than com-
plete interrelationships of all the pieces. In No. 57B, for
instance, questions of what happened to the White force
are needed to disprove the retrograde try Pf6 X R or S.
But especially interesting is the try Pc7--c6. If we
replace this we see that the Ba7 has been bottled up in
an imaginary district (see p. 13) from which he can never
· · · return to his original post at f8. •
No. 58 is a fine example of the complete analytic of
''
: i the side-square type, and it contains a pretty artifice,
which will reappear with great effect in later chapters,
WHITE.
13 in the try Ph7 X g6. On inspection it will be found that
the only available White man to be captured at g6 is the
missing R, and if he reappears on g6 he immediately gives
Mate in two.
an irreal check, since there is no 'Black man whom he could
have captured there on his previous tum.

120 121
 No. 59. No. 59.
N. HOEG. I PxP e.p.
Original. Four of the five missing ,White men were taken on
b6, b4, c5 and d6. The fifth, the KB, must have fallen
on a White square, hence the tries Pf4 X g3 and Ph4 X g3
are easily disposed of. Nor can Black have played Ke8-
e7, Pe7 X d8=R, for the P's on a7, c5 and g4 have made
five captures, accounting for all of Black's lost men except
the QB, who never left c8. The Pg4 must have captured,
else the Pg3 could not have got there. It is a pity space
prevents a full study of each of these complex diagrams.
In No. 59A note the retrograde pin of the Black K, as in
No. 57. No. 59B prettily combines an en passant variation
with its thematic key.

No. 59A. Ko. 59B.

R. L'Hermet. \\·. Hundsdorfer. ,
Riga Tageblatt, 1894. Woclzensclzach, 1907.
BLACK.

WHITE.

Mate in two.

WHITE. "\VHITE.
- l\Iate in one. Self-mate in two.
I PxP e.p. 1 PX P e.p., Bfz+; 2 Pd4+.

122
123
 ..
No. 60.

W. HUNDSDORFER.

No. 60.
I PxP e.p.; 2 Pg7; 3 Rh4; 4 P=S.

Count Arnold von Pongracz, the Hermit of Tirnau,

in I863, was probably the first to use the en passant key
in a self-mate, but his position was a very crude· one com-
pared to so skilful a problem as No. 60. Here the merit
of the solution forms an unusual adjunct to the merit
of the analysis. The vital point in the analysis is the
presence of the P at a5. The White captures prevent this
P having come from a7, hence it must have played Pb6 XS
at an earlier stage, and the loss of this \Vhite S disposes
of the tries Rx g2 or Pf6 X g5. The ideal analytic com-
. __ bines, with the interest of its past history, some actual
merit of solution such as is presented here ; but. every ideal
is rare of attainment and almost invariably one or other
of these elements, the past or the future, is subordinated
to the other. Nos. 61-62 have been selected primarily
for the interest of their mates, and with the exception of
No. 6IA the analytic side has suffered. No. 59B combined

WHITE. / s, direct and retrograde play very agreeably. It is easy to

see· that Black has just played Pb7-b5, for Ph4 X g3 is
out of the question, since either White's KR P or Black's
Self-mate in four. KR P must have queened.

124 125
 No. 61. No. 61.

\V. I. KENNARD.
I PxP e.p., Pg5; 2 Kc4.
Pg6; 2 Pc7.
Ill. London News, 27 Dec., 1890. Analytically this closely parallels the elementary
simplicity of No. 55B, but the happy addition of the Pg7
BLACK. produces an artistic and original result in the blending
of an ambuscade with a minor promotion in the play.
No. 6iA leads the two-movers in respect to the number of
variations, with analysis also prominent again. No. 6m
is one of those rare symmetricals whose keys are quite
unsymmetrical (see, for instance, Group 6 of The White
King). Black's last move cannot have been made by the
K P, as the K B must have been slain long ago. The
problem must not be confused with those in Group 98,
illustrating partial e.p. analysis.

No. 61A. Ko 6m.

J. C. J. Wainwright. T. R. Dawson.
Boston Post, 1901. Falkirk Herald, 17 June. 1914.
BLACK. BLACK.
G

WHITE.

Mate in three.

WHITE.
Mate in two. :VIate in two. · _ (',_:./)
l PxP e.p. 1 SPx P e.p.

I:Z6 127 '

 No. 62.

R. BEHAL.
. No. 62.
Neues Ill. Blatt Ty., 1897.

BLACK.
r PxP e.p., Kxg3; 2 Qgr +.
Kxg5; 2 Qg7+.
KxB; 2 Qe5.

_ Of all the e.p. keys, none leads to· as pretty variations

as No. 62. The solution, indeed, contains a distinct and
clever theme, which has been used in at least three other
problems, one of them being No. 89 of Tasks and Echoes ·
and the remarkable thing is that Ko. 62 is the predecesso~
of all the others. It seems almost absurd that the com-
poser of a new theme should deck it out with such an
unorthodox key. Such a thing is rather to be expected
~fter _th: t~eme has become hackneyed. And a further
mcongrmty 1s that the problem was entered in an orthodox
. __tournef, _where its ~hances of recognition were nil, nnless
?Ur op1m~:m about Judges and analytics given on p. rr7
ls very wide of the mark indeed.
. Speaking of e.p. keys leading to co~tents noteworthy
m themselves, le! us refer again to Wolf's No. 24B in Tasks
and Echoes, which shows how a whole horizontal line
can be cleared of obstructions in a three-mover by using
the e.p: key. ¾et us ~efer also to No. I5B of the same book,
WHITE. by Erhn an_d ~emo, m which the capture is repeated on
each of White s three moves. That problem is very well
Mate in three. constru~ted. analytically, requiring a complete study of
the White side to show that the Black K is really in retro-
grade pin. ·

' 128 r29

J

[
- - - - - - - - - - . . . - . , .-'\,,,..-_- :-.,..,,...'l'._ _ _..
J - - - -

·l.

No. 63. No. 63.

W. HUNDSDORFER. I PxP e.p.
Original. -Heretofore in every conventional example, the theme
P was compelled to move by a retrograde stalemate of
BLACK. Black. A very pretty advance on that motive is for the
retrograde stalemate to be transferred to White. This
requires the theme P to move to release White, and the
'\Vhite K on the side square forces the theme P back to
its original square. Thus, in No. 63, Black has apparently
many last moves at command. Closer inspection shows
• that the only legal release of stalemate is by r Pe7-e5,
Kd4-d5 ; 2 Pc6-c5. No. 63A is solved in retrogression
b}' r Pb7-b5, Pd2-d3; 2 Sb5-a3, Rar-a4; and
No. 63B by I Pb7-b5, Sbr-a3; 2 Ba--t--e8+, cleverly
anticipating an imaginary check from the Raz.

Xo. 63A.
T. R. Dawson. T. R. Dawson.
Original. Original.
BLACK.

WHITE.

Mate in two.

WHITE.
Mate in one. Mate in two.
I PxP e.p. I PxP e.p.
130
131

"- ' ._,.,-_, ,.,

. .,. -". . -- ' ., . , - ,. ·g s -~ ~
 No. 64.

A. TROITZKY. ,
No. 64.
V., Magyar Sakkitjsag, April, 1913.

I PxP e.p.+.
The essential idea in No. 64 is that the White K shall
be driven at some time in the retrograde development to
a side square: if the theme P had stood on the passage
square he would then put the White K in imaginary check.
It is in this way that we obtain the conventional type in
its most beautiful form. Here there is no longer an immedi-
ate imaginary check if the theme P retracts to the passage
square ; the thematic effect is mttch heightened by having
this now familiar feature deferred to a later stage of the
retrogression. In No. 64 the theme P m11st retreat to give
White a move. If he retracts I Pb6-b5, the retrograde
play continues Ka4-a3; z Rb3-b4+. Now, owing to
·-·the theme P's guard on the side s::inare a5, the \,Vhite,K can
ne,·er move to that square. He is, perforce, driven back
to a3, in turn forcing the Black R back to b4, and so
perpetital retrogression makes its appearance, a new irreal
phenomenon in our retrogressive analysis (see p. 13). \,Ve 1\i_,.ii\'·
define this condition of perpetual retrogression as any .
stage in a retrograde sequence of moves where further
moves may be added to the sequence, which moves, how-
ever, always repeat themselves, either immediately or in
a cycle. That the phenomenon is irreal in all cases is
Self-mate in one. obvious. It is most interesting because of its analogy to
perpetual check in direct chess, and it is very fruitful of
deep retrograde tries; see Nos. 79, 79B, 83B, 87, 92, 95B.
 No. 65.
No. 65.

\V. HUNDSI?ORFER.
, I PXP e.p.+.

The essential idea or Groups 65-67 is that Black shall

Original. be in retrograde stalemate except the theme P ; this, on•
retracting, exposes the Black K to check and in abolishing
BLACK. this check some piece must sooner or later travel across
the passage square. No. 65A will easily be understood.
After the theme P has retracted the check of the Bb4 has
to be accounted for. It could be given only by Rd6-b6 +,
which move carries the White R across the_ passage square,
c6, requiring that square to be unobstructed. In Ko.
65B the crossing of the passage square is deferred, thus :
I Pe7-e5, Sf6-g8 + ; 2 Bg8-d5 +, Sd7 x B + ; and in
Ko. 65 it is again deferred to the fourth retracted move:
1 Pb7-b5, Rc6-c5 + ; 2 Sd6-e4 +, Bd4-a7 +.

:Ko. 65A. Ko. 65B.

A. \Vaterhouse. E. Eginton .
. __ Birmingham Post, 27 June, 1913. Gazette Times, 31 Aug., 1913.
BLACK. BLACK.

W'HITE.

Mate in two.

WHITE. WHITE.
Mate in two. Mate in hvo.
·l PxPe.p.+. l PxP e.p. +.
 No. 66.

T. R. DAWSON. No. 66.

f
Birmingham Post, 23 June, 1914. l PxP e.p.+.
i
BLACK. ,\ Prior to the last group we aiways deduced the absence
I
of the theme P from the passage square because of the
imaginary attack on the ·side square. We have now
entered on an investigation of the retrograde obstructions
or blocks of the passage square, which furnish a totally
new means of showing that the theme P cannot stand on
that square. Chronologically, Loyd was the first composer
to break the spell of the conventional type, with );°o. 77,
dating from 1894. There was another example in 1898,
No. 77A. But the real birth of the retrograde study of
legal unconventional en passant keys occurred in 1907-8,
when A. Troitzky, A. C. White, W. Hundsdorfer, and
A. Batori simultaneously turned their attention to the _
S'!Ioject-:-- (See Nlemories of my Chess Board, p. ror.) Their X
... early efforts were rapidly elaborated. The device of
retrograde tries is one of the extensions· of the .simpler
themes which we shall frequently meet. The- addition of
false retrograde lines has the same pleasing effect upon a
-solver that has the weaving of harmonious variations in
a good direct-mate. When the false lines run to two or
three moves, before becoming irreal, the effect is very
WHITE. rich. The theme of No. 6_6 is the simple one of the Rd6
crossing the passage square. A fine try occurs by retracting
Mate in two. 1 Pe6-e5, Kf6-g6 + ; 2 Ph7-h5, and now no matter
what retraction we make for White, Black stands after it
-in retrograde stalemate. Note especially the Sg3. · With
the White K at g6 this S is held in a retrograde pin. With
K at f6 the S's moves offer imaginary checks.

1 37 '
 No. 67. No. 67.
I PxP e.p.+; z Qgr+; 3 Se3+;
W. HUNDSDORFER. 4 Sgz + , Kcz ; 5 Bg6 + ; 6 Qc5 + .
Original. There is one special division of this cross-check theme,
in which retrograde tries 'are a necessary part of the play.
BLACK. We refer to the case where the White piece which checks·
the Black K is the same piece which crosses the passage
square. It is almost needless to add that this White piece
can only be the Q. In No. 67, if we replace Pf6-f5, we
obstruct the passage square to the \Vhite Q. She can only
abolish check by retracting Qf7-g6, whereupon we find
Black in retrograde stalemate. If the theme P returns
to £7, the Q can travel over £6 to a6, unchecking and suit-
ably releasing Black. Xos. 67A-B illustrate the same
pretty thematic tries.

Ko. 67B.
"\Y. Hundsdorfer. T. R. Dawson ..
Original. Bolton Football Field, 28 Dec.,
1912.
BLACK. BLACK.
==== :::;===;:::;;;;;;;;,-,
';~ ,~:
:<~

WHITE.

Self-mate in sii.

'· WHITE.
'L ::.\Iate in three. Mate in two.
1 PxP e.p.+.
w. HUNDSDORFER, after A. TROITZKY. , No. 68.
Original._
I PxP e.p.+.
BLACK.
In Groups 65-67 the obstruction of the passage square
by the theme P depended on Black being in retrograde
stalemate. We shall find, in Groups 68-71, that a similar
obstruction turns upon \Vhite being stalemated. We
divide this material, which contains some of the prettiest
and most difficult in the whole work, into three sections
depending .on the checking motives involved. The essen-
tial idea of No. 68 is that the White ·K is forced back into
a real check and that the theme P must leave the passage
square clear in order that Black may abolish this check.
Owing to ·the White retrograde stalemate the theme P
niust return to g6 or g7. Previous to that the \Vhite K
must have stood on g5 ; and now the check from the Rd5
. can only be accounted for by the Bh7 having ·crossed the
-· passage square. It is worth note that the check'of the
\Vhite Q has nothing to do with the theme. Even if the
\Vhite Q were a R, the K would still have to retract to
g5. The problem by Troitzky on which this composition
is founded will be seen overleaf in Ko. 69. :rt speaks of
the difficulty of classification that themes so closely allied
must yet be separated. This theme can be carried deeper
WHITE.
into the retrograde development and the passage square
crossed by Black on his third or even fourth last move
Mate in three. instead of on the second last move, but as yet such demon-.
strations have only been made in incomplete or unorthodox
form (see. p. 89) .

..,
 1
No. 69. I'
'·
No. 69.
I PxP e.p.+.
A. TROITZKY. ...
. Instea~ of a re~rograde check actually occurring and
Original. bemg abolished, as m No. 68, an irreal check, which never
occurs at all, can_ be equall):' fu1:1damental in determining.
the th~me when its prevent10n is the thematic necessity.
:, . Thus, m No. 69, after re!racting r Pg7-g5, Rg5-g4+ ;
we obse:ve ~hat ~he White R is held in retrograde pin,
threatenmg ima~nary check if it moves again. Hence
2 Rf6--h6,_ crossmg the passage square, is a further neces-
sary rel~asmg move_. No. 69A goes deeper. Sooner or later
the_Whi~e K must go to g;7. But the Rf7 would then give
an imagmary check, and it becomes uro-ent to release him
thus : I _Pb7-b5, P):>5 X R ; 2 Re6-a6,°Kg7-h8 ; 3 Rf8_'..
£7 +. :No. 69n, which interchanges the duties of Black R.
a!1d B as compared with Ko. 68, was crowded into the
present page through space _requirements.
Ko. 69A. . Xo. 69B.
\V. Hundsdorfer. \\·. Hundsdorfer.
Original. Woclzensclzaclz Ty., 3 Oct., 1909.
1
:!
==B=I.=AC=K=. ===::::-i''t =~;;,:='=:, ===• l1_,
l.
'.1

WHITE.

1'

l\fate in two. -1

\VHITE. 4 WHITE.
· 13
Mate in three. f
:\!ate in .three.
I Px p e.p., Pf5+; 2 Bb2+
PxB; 2 RxP+ .. I PX P e.p., Bh7+ ; 2 Bf5+,
 -- -· --------,·-·---..... ~
.................._

No. 70.

No. 70.
I PxP e.p.; 2 PxS or Ph7.

N on-check:in~ _themes '7J!

. White in retrograde stale-
to whether it is White or
mate may ~e di':'1ded aico forc!d back across the passage
Black who is gomg t<;>d e Black :first. Here Black must
square. Let us con:si e~ece so situated that Black shall
uncapture some White p e uare before he can do
be forced back across the. passaf t~s extremely difficult
it legally. The mechamsr 11o The White force save
t heme is brought about as o o~s.f tened up by suitable
one piece. an.d an odd p or
t so critical
is as moment
. the White
blocking de~·1ces. Thei1!ckap which releases some 1?l~ck ,,
piece is sacrificed to a . b th p During the remammg ;
piece hitherto blocked_ in y e k with the odd P's left .
fe~ moves which White ;11ay_ m~li~ed to cross the passa?e '-l.__
him, this released Bla~ piece is o . ite illustration of _t?l-s, <..,
s uare. Ko. 70 conta1~s ~n exqmsro rade demonstration. ~ ~
· i,','volvh,g legal uncasthng '~Jh:1:0n~ ;,. the White K
The Gapture here was_ Pdi,X ;_ the K side to move. )Vh_ile f
'"'d ·~
leaving only two "'.b~te : ~5 and h6 it was essenti~l to , ~" '-
these P's were arnvmg a f 11 ed through the time- ---.;
bring out the Black, QB, \~QR
who, after crossing ;:- •
saving device of Ca~thng,. bleJ at h6. , In order to prev:ent v;i>
the passage square, is sacn:6. ·h. e beautiful conception, ,,
any failure to understand t is v ry y well be repeated : '>
the exact sequence of d~t6m~g4 X Q . 3 Rd8-d6, ,'.::
, I Pg7~g5, Pg5 X R ; 2 R .
. Pg3-g4 ; 4 C;:i.st~, Pg4-15 R + Tableau ! . . X
.../ ,_
5
Qh~.h5 +' Pf3 X B ; :.:.::-

~
6 Bc8-g4, Kb,~s ; 7 P 7
R. ,·.
1
., .-
~ ~l
1/,••v
VI v',.. o L,.,,._· '
r
• ,
r!
'7 I I IA /1,
~· 1 ? c , .
V, (. v / \ -
Ti, ·. ,h U
. .
1

•._/.",.,,,,,,.✓,I(, :,r
/ . 1P . 1
·' .~ . f ~/'Iv
k/V➔ I ~ - ~
v-( .
): Cc.vrk-1.,_vu.- t~--:--· ·. · .· . · 1
.

. . .. I ' 0. /)
/.·. w,
. :6 'f.4,., i
,.·
?c. G J.,, J.u,a.-·f{i"',
·1-. •
~ J) t-.
 No. 71. No. 71.
T. R. DAWSON. .I PxP e.p.+; 2 PXR=Q+.

In _th~ cas~ of. White crossing the passage square, the

Stratford Express, 20 Dec., 1913.
essential ~dea is ~hat, after the theme P has released White
from an impendmg retrograde stalemate, the released man
BI,ACK.
1:1ust at once _traverse the passage square. No. 71 is a
simple expression. After we replace I Pb7-b5
that the White Q alone is free and must have 0 s:i
from ~6 or somewhere further 0;1 the sixth rank, following
co~:;
Black s check: 2. Rb5~a5, which the withdrawal of the
the1;1e P has mcidentally made legal. No. 71A utilises a
White ~ and No. 7m a White B to cross the passage
squa~e, m both c8:ses Black's previous check being again
legalised by the withdrawal of the Pawn.

Xo. 71A. Ko. 7IB.

T. R. Dawson.
. •· V. Bolton Football Field, 27 Dec., T. R. Dawson .
1913.
BI.ACK.

WHITE.
'I
Mate in three.

\VHlTE.
])late in two. Mate in two.
I PxP e.p.+. I PxPe.p.+.

1 47
 No: 72.
W. HUNDSDORFER, after A. BA!ORI.
No. 72.
Tidskrift, 1908. f I PxP e.p.
BI,ACK. The present division of the classification, Nos. 72-76,
really presents only" a special case of the last, but the
mechanism of the various themes has a distinctive character
deserving of special treatment. In this instance the
theme P clears both the present square and the passage
square, instead of only the passage square; so that the
file shall be open either for Q or R to move up to the passage
square, or for an opposing P to regain the passage square
with an open file below it. The material falls into three
sections, Nos. 72-3, 74, 75-6, parallel with the previous
sub-division in Nos. 68, 69, 70-1. Thus we first take up
the case of clearing for a piece which would otherwise
give an imaginary check. The idea was first propounded
by A. Batori, and logically simplified in No. 72. Here,
White is in retrograde stalemate and the only way to give
him a move is by replacing Pd7-d5, clearing the file for
-- Qd6-:--d.2.
Some of our readers will surely have noticed the analogy
which the passage square in all this retrograde analysis
bears to the so-called critical square in the direct chess
of the German interceptional school. Space prevents any
WHI'.rE. accurate following up of the analogy, but we give the hint
for what it is worth to those interested, hoping that someone
Mate in two. else may have· the opportunity to elaborate it. Suffice
it to add that Groups 65-71 bear the same relationship
to Groups 72-76 that Loveday's Indian bore to Healey's
Bristol.
 No. 73.
No. 73.
T. R.
. DAWSON.
. I PxP e.p.+.
No. 73A transfers the theme of No. 72 from the White
Magyar Sakkvilag Ty., 1914. Q to the Black Q. The clearance I Pc7-c5 is necessary
so that after Kb4-a5 the Black Q can abolish what would
BLACK. -r otherwise be an imaginary check by 2 Qc6-C4 +. No. •
73B is an economical setting of the direct clearance for a
White Rook, thus : I Pb7-b5, Rb6-b4 ; 2 Kb5-a5,
Rc6-b6 + . The clearance may have its purpose deferred
by a pretty artifice, as in No. 73. After I Pe7--e5 has
cleared, we do not see the reason for it until _after Re4-
g4 ; 2 Rg4-h4 +, Re6-e4 ! The more the purpose of
a retrograde clearance or other strategical move can be
deferred the more subtle its effect will be, just as is the case
with deferred strategical manceuvres in direct chess.

No. 73A. Xo 73B.

T. R. Dawson. T. R. Dawson.
Hampstead Express, 26 Dec.,
Gazette Times, 21 Dec., 1913. 1914.
BLACK.
r.=~== ====;:"•

.i

WHITE.

]\fate in two.

WHITE.
l\Iate in two.
I ~XP e.p. +, Q X P ; 2 B X
, Q+, Kc8; 3 P=Q+. I PxPe.p.+.
 No. 74.·

T. R. DAWSON.
No. 74.
Reading Observer, 20 Dec., 1913.
Coming to the indirect checking clearance theme we
find the essential idea to be that an imaginary check is
threatened on one or the other King, and that the theme P
must clear the file in order to let some piece cut off the
check. The conception is closely connected with that of
No. 69, the difference being that instead of traversing
the passage square the relieving piece uses it for a vertical
clearance. The theme may be doubled, as in No. 74,
where two Black pieces are needed to free the Sfz, which
is held in a double retrograde pin, imaginary check being
threatened to both K's if the S moves. To relieve this
pressure Pb5 clears to b7 so that there may occur Pb3 X B ;
2 Bc6-a4, Pcz X R ; 3 Rb6-b3, Pa4-a5 ; 4 Bg2--c6,
Pa3-a{ ; 5 Rf6-b6, S-fa ; 6 Rf2-f6 +. The capture
on the left wing of both the relieving pieces greatly
intensifies the difficulty of the theme. Strictly speaking
this problem shows not only an imaginary checking theme
but also a real check. The imaginary check ·on the Black
King may, however, be said to determine the theme, and
as examples are extremely rare it was hardly woith. while
to make a separate section for the double theme in which
the real check on the White King was considered. So
difficult is the theme, indeed, that no orthodox rendering
WHftE.
(see p. 89) has been composed, and so, the one solitary
instance in this book, we have made an exception and
En passant is legal:. given a pure analytical study to illustrate the point in-
I volved. It is remarkable that a resort to this compromise
.l was required only once in the whole classification.
1
) I.
11
Al

~· 't
..
•,<

. .
.:L;.;;,_.~x,-,:,w,,,,,.:,,,,·-•.f'~·:..,;~,·•,.;-,-;f4i.f . •. 9s.,>-'Y,;.,-<;,."'.•"'¼.•,l\h\•·,;,,,~M·•s-'i,l#:l!fil"':'➔~;;s,•..•.. ~'t.P·,~ ... ,·.·, .·,,. •·
I
l
No. 75.
I
No. 75.
PxP e.p.+, PxB; 2 KxP+.
i' W. HUNDSDORFER.
f
t The non-checking clearance themes are parallel in
I Pr. Set, TVochenschach, 1909. every way to those of Nos. 70-71. It is·simply a question
of relieving threatened retrograde stalemate by giving
BI.ACK. back a piece which was captured at a crucial moment in
the play. Thus in No. 75, which with _Nos. 80 and Sm
formed the splendid set awarded First Prize in the en
passant tourney of the W ochenschach, the crucial capture
was Pg7 X B, which left White with only seven P moves
(KS P and KR P). During these seven moves Black had
to bring his l-{. Rand KB, just released, to cz and a7. In
the retrograde sense, therefore, Black has to clear I Pc7-
c5 to permit Ph4-h5 ; 2 Rc6-c2, Pg5-g6 ; 3 Rg6-c6,
Ph3~h4; 4 Rg8.:._g6, Ph2-h3; 5 Bd4-a7, Pg4-g5;
6 Bg7-d-1-, Pg3-g4; 7 Bf8-g7, Pg2-g3 ; -8 Pg7 X B !
~O. 75A. No. 75B.
\\-. Hundsdorfer. A. Batori.
Allgemeiner Zeitung, 1908. Original.
BI.ACK. t.q-

WHITE.

WHITE.
.)late in three. I/ Mate in two .
I PxPe.p.,BxP; 2 Bg5. I PxPe.p.
 ....,·. -
No. 76.
Original. U ncastle, and play I PX P e.p. +.
BI,ACK, Besides the very difficult type of non-checking clearance
themes illustrated in No. 75, there is a large division of
clearanc~s for an opposing P (see p. r49) which strictly
belong to this section. The essential idea therein is simply
to withdraw the theme P so that a White P on the seventh
rank may uncapture on to the passage square and find
a clear file behind it. For example, in No. 75A, Pf5 clears
to f7 to permit Pf6 X R, the White Pf6 being now able to
return to fa without trouble. No. 75B is another setting
of this simple theme. In Ko. 76, Troitzky has succeeded
in extracting from this arid material a most exquisite
point. As in No. 29, the p9sition demonstrates analyti-
cally that White has just castled. Now since this move
has actually occurred it follows that neither the White K A _ 1 • '·
nor the Q_R has previo~1sly moved, and this _point II;'.11St ---:t'-'- ?" ''
be taken 1nto account 1n the further analysis. jVh1te's /,-,,~
· .. previous move could not have been Ph2-h3,_because the /r, }
Black KP must have promoted via h2.------Nor could it· ' rv 1 ft''-'.,
have been Pc6x R, because the White QB P must have v:c h«J/
promoted.s_ Nor could it have been B-£8, as this would' ,.s'
leave Black:i in imaginary check. Nor could it have been
WHITE. Bb3-a4, followed by Black's Pqxb3, because Black
.made this'capture long ago to let the White QB P pass .
White retracts his fast move and mates in three. . The onJy other move is Pa6 X R, ~nd this would be illegal
so long.as Black's QR P remained on a5, as the Pawns
. a5,/26 could not have passed. Hence the retrograde.

t
I
I
~uence : I Pa7-a5, Pa6 X R ; 2 Bc5-b4, Pa5-a6 ;
3 Pb4-b3 ! . . . . .
! .. j ., , . A,,: ..1 1~.,--· .\,i~-J ,,:.~_;,Ct.-(,:,, t. .,,, , r',) '~'c, I ,·.,.'5.. ;f:)_.:---
VV'--V'--' r'" . ~<..- ,..., , , - I I , ' • I
, • • I i - r \ ., , i ",,
11,g r n j•·e,,vy-f ,:,»-, r' J \, <J,.·•vJ .,._,, -(. 0 l-.C-0··-
--t
. .(, -~ · . . :.
I";"'
~ r',I .
...._
·-<. I-: ~1 " " [ : , ' u)_. \ 11,.
""\
.I
:"'\!-( s.,_,., ,.-0 \,-l.·

(r
h 157 . .•,

1<--;P,, ( 0~;_,';J-o r}/ ~~ Jv~~;z:f

✓ -...~•-;'c ,J /---/- ----!.7,----;:~f, LU,;.,,-:. e ~ ~

~~--=-.. . .--~~-~,,;,,,,,,=-~""""'=~,,,,,,,,~-~~=-~•
I?,
v~"<e ~ - e rf,e/1,<>yi~'"tot.Arf",.,,;:, t?...-<1-<. ,...,,,,,,.~
.n .
i: -, .• '
: ~. .C,,-i,vt. 1,,' ,./.,.,,,.,••.';-;,f._.t, J.,,..f,-~ , v }'1-:- 'ft-1..-/.v"/~l..1~..t.,r .,;/;4 ,J,-,,,,W- t ,
~"'°•·"""·""'·•··,;,,_,.,,,,.,.,_,_,_,,,,,,,,,,J""·,·"'"••-=.,.,:=,.,,""'-'K"""·.-""'-..'·""·""-•"""''·;,..,.,..,.,.,,,,,_,.,,.,_;,_.,,.,.,_"""""..,,;.,_~---·""''"'""'···""'·-,•·"",-,.'~.,,,,..,._
- ..,,, . ,.,,,.,,.,....,,.,
..,.,,,,
....,,,_
... """""~~ -
 No. 77 .
. No. 77.
I PxP e.p.+; 2 Rg5.
·S. LOYD.
We have now proved the absence of the theme P from
the passage square for three reasons, the attack on the
side square, the obstruction of a rank or diagonal contain-
ing the passage square and the clearance of the theme
file. We here enter on a new set of arguments concerned
no longer with lines of action but with the simple blocking
of individual squares. It will be found that the dividing
line between blocking and obstruction is often indefinite.
We have therefore adopt~d the rule of thumb that if the
piece concerned with the passage square could rest upon
it, then the problem has a block theme ; put if the piece
in question is forced to cross the passage square and must
rest further along its line of march, then the problem has
an obstruction theme.

No. 77A. No. 77B.

K. von Ungern Sternberg. A. Troitzky.
Diina Zeitung, 1898. Schachzeitung, 1907.
BLACK.

Mate in three. :Mate in three.

l. PxPe.p., BxB+; '2 RxB. l PxPe.p., 'PxR; 2 Pf7.

159
 No. 78.
No. 78.
T. R. DAWSON.
I PxP e.p.
Original.
We now meet the case where both sides are almost
BLACK. i5 stalemated in the retrograde sense. Loyd's No. 77 was
the first en passant analytic to break away from the
conventional side square type (see Loyd's book, p. n5).
Here Black has all his force in retrograde stalemate save
the theme P. If he retract this to f6, he blocks the White
R, and now White is in complete retrograde stalemate.
Thus the theme P is forced back to f7. This fruitful idea
strangely enough was overlooked on the Continent for
several years. In 1897 Seyboth raised the question
whether it was possible for an en passant key to capture
a Rook's P. Of course the side square type could no longer
be used for this purpose. The first correct answer in
problematic form was No. 77A, though Seyboth himself
had composed an earlier illustrative sketch. No. 7JB is
I
a curious hybrid which shows the obstruction of the'passage
!
square to the Re6 and its block to the Bh4, · an excellent
example to point the distinction between obstrucfion and
block. An interesting elaboration, first developed by
j\ ,! Troitzky in No. 79A, occurs where the threat of an impend-
ing double retrograde stalemate can be deferred one or
I.:
'''
\VHI'l'E.
I more moves. No. 78 is a task record in this regard. It
shows a simple block of the passage square to the Sa8 ;
but note the imaginary variation : I Pb6-b5, Bb7-c8 ;
2 Bc8-d7, Rd7-e7; 3 Re7-e6, Pe6_xS; 4 Sh6-f7
Mate in two.
producing retrograde stalemate of White.· If, in this
sequence, White retracts 2 Rd7-d8, Black is stalemated ;
or if 'Black retracts 3 Be7-f8 White is again stalemated.

161
 >. 1/'
~1i
0.
lu;1
No. 79. ./ .
1v ::,.. l .. -.

t; ,,>) /;r .~ · ·· :·. /,._t"')t >1::

PxP e.p.+.
I ~ .
.
. I •~,. . I.I
I The imaginary tries of Nos. 78 and 79A remind us of
· the tries in Loyd?s I4-I5 Puzzle (Loyd Book, p. 99). An
, j even more interesting extension is the introduction of
~ ~ t u a l re~~sio!_l (p. I33)._ In ~o. 79B, after I Pc6-
.' c5, B'q=o"'S; 2 Qb8'-a8, BdS:-q, 1t becomes useless to
look for any further possibilities, the· pieces which are
apparently free ·continuing to oscillate backwards and
forwards like caged animals. No.· 79 increases the range
over which two Bishops may wander perpetually if the
-theme P only retracts to c6, for the squares a7, b8, q, d8,
e7, £8 form an area froni which they can never be released.
These monotonous sequences could only be interrupted
by allowing one side to make two moves in succession,
clearly an illegal expedient.

:No. 79A. •
A. Troitzky. .
 No. 80.

W. HUNDSDORFER.

I Pr. Set, Wochenscliach, 1909. No. 80.

BI,ACK. I PxP e.p.+, BxP; 2 Sg5+.
In all the problems of Groups 77-79 the composer,
having" a double retrograde stalemate ever before him, was·
forced to contrive some interlocking of the forces, usually
very obvious to the solver, though most refractory to the
composer, especially as to obtaining any orthodox mating
stipulation. To lessen this drawback: we may bring the
mechanism of the theme P's double move into connection
with previous play by holding one side only in retrograde
stalemate, forcing the opponent to release the constrained
side. We have already seen much use of this principle
in the chapters on obstruction, and the present examples
will be closely parallel. In the non-checking themes the
. _. retrograde stalemate must threaten the White fore€, but
the blo.± of the passage square can be arranged to hamper
the Black men equally well with the White. The next
group will show the release of the Black men. In No. So
the White Sf8 would be blocked in if Black retracted
r Pg6--'-g5. In order to release White completely it is
necessary to play : I Pg7-g5, Pg2 XS ; 2 Sg5-h3,
WHITE.
Pa2-a3 ; 3 Se6-g5, Sg6-f8 ; 4 Sf8-e6 +.
Turning to the block of the Black men, we come to the
most intricate and difficult of the blocking themes. The
Self-mate in two." essential idea is that the theme P and some other Black
man must both move in order to release an impending
White stalemate, and that the theme P must not block
the passage square to the other Black man. ·

"
 No. 81. ·
) No. 81.
T .
Ip .
xP e.p.+
A. ~OITZKY. h.mges
he was
crucial
Ph capture on which . .
Original.
~to~vcly f{ ':''t mo:!': a
5 Ph3~h• 3 Qh1-c6 0
~'. six ~gular positU>n
P 7-c5, PqxQts' _and reached
2, Pd -d ' c3-C4 · , 2 Qc6-b
move cannot h3 4; 6 I'h4 h• 4 Ph2-Q p 5,
sh
R p trnvell d ave ~een I
f3. No, e 0 ~ h3 at full
3
3, Ph3 x B. Jl1 c2,-c3 ;
f3---gr °'
ack ' last ~
~
blocked J,ould ,t have been s~eed, as did the' the Black
Pcs X R ;
s,ngle ste
Black Q. In Jo
c6-c5, " that
Rc6--b6, Bb' . 8IA we xetrnct ould have
Q OVeI ';.ew
m · P, I Pc6-- v--.,5. Th I PC']
a ~';;,~ to release th?B would block ~:etrncHon of c5~
Ra . B, thus, I Pa a5. No. 8m sh0 Black R from
b/1\i-='.,~6, Pb,_!b ~5, P<4xB; ,
3' 6 Pb7xB '3 e4 crucial
Rb4~a4,th Pc3-
6tbt block of
- ., 5,5 Rb8-
Pb3x
i!
captu c4
No. Su. re.
T. RD
Gazette r·iines,. awson.
21 ~larch
BLACK ' 1915.

~) .
p,;,

WHITE. &,_

WHITE
:Mate inf
. .our
I PxP e.p., BxP.' .2 S xB.
166 •.,
167
 ' t

No. 82.

W. HUNDSDORFER.
No. 82.
Original.
I PxP e.p.+.
BLACK.
.Cross-check themes permit the retrograde stalemate
to threaten Black (see Groups 65-67) as well as White.
Ko. 82 is a fine, though not difficult, example of passage
square blocking, with Black in retrograde stalemate. -
The last moves have been : . I Pd7-d5, Pd5 X R +, leaving
\Vhite in an easily explained retrograde check, which
would become imaginary if the passage square d6 were
blocked against the Black R. The close thematic relation-
ship between square blocking and line obstruction is well
shown by comparing No. 82 with Ko. 65B. In the latter
it was necessary for the Black B to cross the passage
square in moving from g8 to d5. Occupying e6 by the
. __theme P, therefore, would have been genuine line oqstruc-
tion. In No. 82 the Black R may have come from b6 or
c6 as well as from d6; but there._is no proof that he came
from beyond d6. By our rule of thumb, therefore (see
p. 159), we would define the occupation of d6 in this
instance as simply. square blocking, though we have no
eYidence that it was not equally line obstruction.
The threatened imaginary check may be deferred one
WHITE. move by being transferred from a blocked Black man to
a blocked White man. Thus, in No. 83A, the theme P
is forced back by Black retrograde_ stalemate.. Then must
Mate in two.
follow : I .. , Rf5-d5 + ; 2 Kd5-e6 +, and now if the
theme P were on f6, the Rf5 check would be imaginary.
I Pf7-'-f5 allows White to retract 2 •. , Rf6-f5 +.

'I
.i

168. 169
 No. 83. No. 83.

w. HUNDSDORFER, after A. TROITZKY. r PxP e.p., Pd5+ .; . z. Kb7+.

r The idea of a retrograde check which would be imaginary
. Original.
if the passage square were blocked can be combined with
the release of White from retrograde stalemate. No. 83
shows the Black R captured. After the theme P retreats,
we have r .. , Pa5 x R, and so c6 must not be blocked.
In No. 83B White is threatened by a~e.tmgression, ,
his K moving on b7, a6, b5, and the Black QB on b5, c6,
b7. To relieve this the theme P must open a path for the
K to b4, but must not block c6, which the Black B requires.
We have now examined every method of proYing the
absence of the theme P ,from the passage square ; and we
will proceed to study its necessary presence on the original
square.

T.
c::~.
hawson.
Xo. 83B.
T. R. Dawson.
V., 2vorwich 11-Iercury, Dec.,
1911 ..
20
Original. 1/
BLACK.

f,

Mate in three.

Mate in two. Mate in one.

1 Px P e.p.

170
 No. 84.
No. 84.
W. HUNDSDORFER.
I PxP e.p.+.
Afiinchener Neiteste Nachrichten, 21 March, 1910.
I
The large variety of motives to which we .are now
BLACK.
proceeding all require the theme P to go back to its original
square at the last move to cover a line, diagonal, horizontal
or vertical, along which would otherwise act sooner or later
an. imaginary check. The first general division, much the
smaller, shows Black in retrograde stalemate. Here the
essential idea is that the theme P, forced to move by the
stalemate, exposes his K to check, and that in abolishing
this check sooner or later it is ilecessary for the original
square to be occupied covering a line. We only quote
one example out of less than ten which exist. · In No. 84,
if we try to retract I Pe6-e5, Pd4-d5 + ; we shall find
the Pe6 held in a retrograde pin, and the whole Black
force stalemated. Consequently the only legal retrograde
play is : I Pe7-e5, Pd4-d5 + ; 2 Be8-f7 +, this move
of the B having become legal through the cover of the
· line b7-g7 by the Pe7.
-· The second general division, with White in retrograde
stalemate, is as usual much the larger, and it can be sub-
divided into two classes, according as the thematic cover-
ing protects the Black or the White K. An elementary
three-mover by A. C. White in the Scha~hzeitung for
December, 1907, was the first to demonstrate this return
WHITE.
of the theme P to cover the line of its original square.
Since then, apart from solitary examples by Batori, Onitiu
Mate in two. and Waterhouse, none but the three masters of retrograde
analysis, Troitzky, Hundsdorfer and Dawson, have experi-
mented with this fertile theme.
 No. 85. No. '85.
.

r PX P e.p. mate:
T. R. DAWSON.
· 'l'~e co:"e:ing of the Black K by the return of the theme
Northern Whig, 25 Dec., 1913. P to its on&1nal square may be necessitated by five specific
· reasons, which we shall t_'.1ke up in turn in Groups 85, 86,
87-8, . 89 and 90 respectively. 'l'he first of these is the
cove:mg to let a White piece occupy' the line. No. 85A
requires r Pb7-b5 to coyer the ~lack K so that White Q
m:3-y move . to _c7. This constitutes horizontal. cover.
Diagonal ~over is s~own in No. 85B, where r Pg7-g5
allows White to contmue _Bbz-~r. 'l'he line-covering may
be_ defer:ed beyond the immediate move, with the usual
heightenu~g of effect. So in No. 85 we have the necessarY
retrogress10n : r Pc7-c5, Pfz-f3; 2 Bf3-h5 Ph2-h3··
3 Bc6-f3, Qe7-e8 ; 4 Be8-c6 +. · ' '
:N°o. 85A. Xo. 85B.
V. Onitiu. .A. Troitzky.
Bi1dapesti Sakkor, 1912. Original.
BLACK.

,vHlTE. i4
Mate in· one.

-WHITE.
Mate in two. Mate in three.
1 PxPe.p.+. 1 Px P e:p:, Rxs+; 2 ·RxR.
 No. 86.

W. HUNDSDORFER.
No. 86.
Tidskrift, 1908.
I PxP e.p., QxR; 2 Rg5+; 3 Pb7+.
BI,ACK.
The second reason for covering the Black K is to unpin
a White man. We have already seen in the Castling
analytics, for instance in No. 50, how, when a man stan~s
a
in retrograde pin, it requires a line cover. to release it.
No. 86 shows how easily the theme P, m en_ P,assant
analysis, can be used for this purpose. As White s last
move must have been R-g2, and as there was no Blac_k
piece which could have been captured on that square, it
follows that the theme P must have been at b7.
It is interesting to notice in how many of these problems,
where a retrograde stalemate is impending, Tthe tur1; _to
move can be accurately demonstrated. In No. 86 White
cannot have moved last, so that no explanati?n of the
position other than that given would be possible, even
if it -were not understood that it is White's turn to pl~y.
· - · Indeed the major part of the retrograde stalemate an8:lytlcs
might have been included in the T?rn-to-mov.e se<:tion of
this work (Nos. 22-25). A ,very little thought will con-
vince one that this is necessarily so. Yet at the first
realisation of the fact there is something rather unexpecte_d
about it which seems to add to the general eccentric
WHITE.
characte; of these analytics. The result is that a com-
poser like Onitiu, first turni~g to _the subject in No: 85A,
_Mate in five. publishes his problem as a tnckwith the colours of his_men
reversed, leaving the solver to_ deduce that ~lack is to
play. For the sake of uniformity we have agam reversed
the colours to make it White's turn to move.

176 II!
177.
 No. 87. ·

Magyar Sakkvilag Ty., 1914.

, I PxP e.p.+.
The third reason for covering the Black K is to unpin
a Black man. By placing the Black man in a checking ·
relation to the White K, the theme P can be employed to
prevent the check from being imaginary. For example,
in No. 87A we play Pb7-b5, so, that ,after K~5-~4 the
Rd5 may be unpinned and the_ Be4 s check not b~ 1magmary.
No. 87 similarly covers the line a7-e7 to unpm the B!:>7·
If we replaced Pd5 only to d6 !a_erpetual retrogress10n
would appear, since the retrogra e development must
continue :· I •. , Kd5~4 ; 2 Sc6-b8 +, Kd4~5 ;
3 S-c6 +, Kd5-d4, etc., forcing the Black S to and from
c6 indefinitely.

No. 87A. No. 87B.

T. R. Dawson. W. Hundsdorfer.
Gazette Times, 21 Dec., 1913. Wochenschach Ty., 3 Oct., 1909.
BLACK. ' BLACK.

WHl'.CE. i.C>

-Mate in four. Mate in four.

I
1 PxP e.p.+, any; 2 BxR. I PxPe.p., Qh2; 2 Pg7.

i~ ~
,~-
W. HUNDSDORFER.
No. 88.
Original.

BLACK.
I PxP e.p., Pf5+; 2 RxB, PxP; 3 BxR+.
Continuing the analysis of the problems in the last
group, we find in ]'.fo. 87B that the Black man to be unpinned
has been captured, a very pretty feature. He reappears
after I Pc7-c5, Pc5 XS ; and clearly if the theme P had
not covered the diagonal a5-d8 the resurrected Sb6
would be in the act of administering an imaginary check.
The problems in Group 87 do not exhaust the possi-
bilities of Line Covering to unpin a Black man. It may
be arranged that the unpinning is required not because of
threatened im:;tginary checks, but because the Black man
in question must move in order to release White from
the impending retrograde stalemate. In this respect the
theme of No. 88 shows a relationship to that of :Ko. 86
· _: not unlike that of No. 86 to No. 85B. In each Gase the
diagonal power of the White B is prevented from being
imaginary by the line covering of the theme P, the B's
power being disguised in various ways. In No. 88- White's
K P died on its own file, as there is no absent Black man
it can have captured: The other six \,Vhit_e pieces and
Pawns were all captured by the P's at dz, f4 and h4.
WHITE. Therefore \,Vhite's last move must have been Pc2-c3,
·,\
j
and his previous move B-a3. So the Black P must have
'Ma~e in four. been home at e7 to allow the White B to come in and ·
C, ()
the Black R to follow to b4- .Each of these ·themes can
be varied by changing the pieces involved, but the results
are thematically the same.

r8q
 - i
No. 89. i
i

T. R. DAWSON. No. 89.-

Original. I PxP e.p., RxS+; 2 SxR.

- .The fourth reason for covering the Black King is to
allow the legal uncapturing of a White piece. No. 89 is
the solitary example of this theme. White, threatened
by a co~ing retrograde stalemate, must be given some
freedom. His K B alone is available and it may reappear
only on b3. Hence I Pf7-f5 covers the diagonal to
permit Paz-a3 ; z Pa4 X B, B-b3.
The fifth and last reason for covering a line with the
theme Pawn for the Black King is for the release of his
· Majesty himself, as shown in the next group. No. 90 is
an ideal retractor; since every move to be replaced may
be established analytically. In this respect it forms a
splendid elaboration of No. JI, and may also be compared
with No. 76. The three last moves must have been:
I Bc8-b7, Pb5-b4; 2 Rb8-b6, the latter reminding us
particularly of Collins. In No. JI the retrograde develop-
ment may be supposed to have been I Rg8 X P, Pg7-g5 ;
2 Ke4-£5, Kg5-h6 ; the theme P covering the' square
g5 and releasing it for the Black K's occupation. So in
No. 90 the retrogression continues, justifying the en
passant key capture : z .. , Pb7_:_b5 ; which covers b6
from the Rb8 for the Black K's occupation. Then comes
3 Pa4-a5, Kb6-c7 ; 4 Pa3-a4, Rc7-d7 ; 5 B-c8,
and the tension is released. That Collins, in No. 31,
though rftaining the conventional side-square arrangement,
should thus have glimpsed the advanced theme of No. 90
is truly a remarkable testimony to his analytic powers. -

182

ro-J'il"'_ ~------~-~-•~"'1'-S!l!.--~L~~--=---=-,-,,"'""-'"-"'"""-""-""'·
m:.!li!O.

~-
.._'"""""•°"'•·'""•._,,~
A""'¥""'«"""-""""•'""''•'""'•"'"'"'"""'•""""'""""'""""'"""""'-"'•-;,,,.,.,
 No. 90. No. 90.

W. HUNDSDORFER. This intricate study, whose so1ution we have just seen

on the last page, introduces us to Nos. 90A-B, which present
Original. a striking idea, combining in a form due entirely to A.
:Muller the en passant capture with uncastling, combina- a
BLACK.
tion we have already met in very different forms in Nos.
70 and 76. In No. 90A I Pc7-c5 covers the square d8
from the Ba5, and after I .. , Pg3-g4 ; permits Black to ·
uncastle without the Black K passing through check on
d8 as he does so. This in turn releases the White K, who
was in· serious danger of retrograde stalemate, as none of
his other P's could have moved owing to the necessity of
allowing Black to promote at dI and hI long ago. Ko.
90B transfers the idea to the King's side.

No. 90A. Xo. 90B. ·

A. Troitzky, after A. l\Iuller. ,v. Hundsdorfer, after A. ::.\Iuller.
Bohemia, 12 Feb., 19II. Original.
BLACK •

. Retract the last three moves, after which the player whose
. turn it is to move can mate. ,..
. --,n ~,
~ ().,A--"
~ . ; l ✓(_ WHITE.
rn.,u S
t _.-
1
• - \ 1• 1 ? ~. -1- .z G 1
....1: ;J. r. Mate in three.
f'1 ')'\,\ 'h (}/'fO r('\/;N f - \. ~, -
f .,,-_
I ·
I · · I Px Pe.p.+, Kh8; 2 QxQ,

€ 7W~ ~ l , ffc.wsi F' '" i %: , .I PxPe.p., RxB+; .2 QxR. RaI; 3 Qg5.

 No. 91.

W. HUNDSDORFER. 1PxPe.p.+; 2Pf7+; 3PxP+; 4QxP+.

We come now to line covering by the theme P on its

Wochenschach Ty., 3 Oct., I909. original square for the benefit of the White K. As in the
'case of the Black K, the possibilities of imaginary checks
BI,ACK.
furnish five reasons to necessitate the retrogression of the
-P. The first reason (see No. 85) is to let a Black piece
occupy the line. This is a troublesome theme to handle.
Nos. 91 and 91A show the Pf5 retreating to f7, so that in
each case the Black Q may go to £6, the White K at fS
being covered by the Pf7. No. 91, strictly spea~in~,
shows Black alone in retrograde stalemate ; but, as 1t 1s
the solitary example of a line cover for Black, in that
state, it would be pedantic to separate it from its own
brother, No. 91A.
Ko. 9IA. Ko. gIB.
\V. Hundsdorfer. W. Hundsdorfer.
Original. Original.
BLACK.

WHITE.

Self-mate in four.

WHITE.
Mate in two. Mate in three.
I PxPe.p.+. I PxPe.p.+.

r86
 No. 92.
\V. HUNDSDORFER. I PxP..e.p., Pe5+; 2 Kg7+.
.r
Original. The second reason for covering the White K (see No.
86) is to unpin a Black man. No. 92 is surely the inost
beautiful problem in the whole of this collection. In it
BI.,ACK.
}3 the German master has illustrated what he calls a .r.etro- ('·
grade vniation. ,'.I'he theme in its elements is that I ~
f-?snall t:60.;r't'he diagonal d5-g8, though it is not till
later that we discover this to be necessary. After the
theme P's retraction White must play Pd6xB or Son e7.
This is where the variations begin. Let us first consider
Pd6 X B. · This must be followed by 2 Bf8-e7. Here we
might interpos~ any number of repetitions o~ the incident. & -r-!tfF.1~
2 .. , Kg7-g8, ·3 Be7-f8+, Kg8-g7+, 4 Bf§.:-e7-;,.M e
\Ve shall only find a real outlet, however, by continuing., 1-,.,;.t1c.p-·-:,.
2 .. , Ph4-h5 ; 3 Pe7-e6 (the unpinned man !), Ph3-h4; V
4 Be6-d5, Ph2-h3 ; 5 Rd5-c5, pfa..:._£3 ;· 6 Rc5-b5,
just in tim~o allow Pb5-b6, whereupon the pressure is
. re1ieved · 1r, ' :•(•,: · . ,
-- Xow let us consider Pd6 x S for White's first retrograde
move. Here again \Vhite has only four P moves left
between him and stalemate, so Black must hurry to his
relief. It is essential to get this restored S to £8, so as to
release the White K from his present pin. The develop-
ment therefore continues : 2 Sf5-e7 +, Ph4-h5 ; 3 Pe7-
e6 (the unpinned man again!), Ph3-h4; 4 Sg7-f5,
Ph2-h3 ; 5 Se6-g7, Pfa-£3 ; 6 Sf8-e6, just in time to
allow Kg7-g8, whereupon the pressure is safely relieved.
Mate in three. The beauty of the problem lies in its wonderful construc-
tion, combining two such deep retrograde lines with a
direct mate in three. No. 9m, also a Black unpin, is
pretty enough but pale before the depths of 92. The last
moves were I Pe7-:-e5, Pc3 x B ; 2 Ba5-b4 +, Ka2-a3 ;
3 Pb4-b3+.

188
 No. 93.

T. R. DAWSON.
No. 93.
Original.
I P xP e.p. mate.
BI,ACK.
The second reason for covering the White K (see No. 87)
is to unpin a White man.· No. 93 is the solitary example.
Pg7-g5 must cover the Black Q, so that the White S may
leave bz without exposing the Kar to an imaginary check.
As we approach the end of our classification, we may
profitably ask ourselves in what new directions originality
in retrograde composition is best to be sought. To many
it will seem that _the complete tabulation of Castling and
en passant themes leaves little to be discovered therein
in the future. This is not necessarily so. Only three
composers can be said, as yet, to have mastered the
technique of retrograde composition and, . wide as the
range of their experiments has been, it is rare in problem
history for the pioneers in any line to exhaust its possibilities.
While they have made originality more diff,icult to attain
they have also furnished materials for the study of the
technique by which these difficulties shall be overcome.
The development of deeper retrograde tries, the subtler
characteristics of perpetual retrogression, the deferred
uncapturing of the opponent's men, the more strategic
WHI'.l'E.
play in the direct stipulations, these and other elements
offer sufficient opportunities to encourage anyone who
feels drawn tq the composition of analytics. In the legal
Mate in one. retractors there is a vast deal yet to be achieved; and
Dawson's No. 26 points the way to an entire new realm;
deep andunfathomed in its possibilities.
 No. 94.

T. R. DAWSON.

Original.
, I
No. 94.
PxPe.p.+ ..

BLACK. Iii . The fourth reas'?n for covering the White K (see No. 89)
1s to allow the legal uncaptming of a Black piece. We
have here an extraordinary theme, very troublesome to
. j deal with. No. 94 is the only specimen extant. The
position is really •a double retrograde stalemate. If we
try I Pg6-g5 as the last move, we may continue : Pc2-
c3 ; 2 Pg7-g6, Pc3-q ; and Black is left in retrograde
stalemat~ ; or we may continue : Pdz x B ; 2 Pg7-g6 +,
and White is left in retrograde stalemate. The onh·
other resource is I Pg7-g5, which covers the diagon~l
a1-h8, and after Pdz XB permits the resurrected Bc3
to slip out of the path of the P now at q.
The fifth and last reason for covering the White K
~see Ko. 90) is to release his Majesty himself, as exemplified
1n the next group ; and it will be found that this final
_s_edi~n of the classification has fitly produced some of the
prettiest of the many interesting problems. •
We may well take this opportunity in parting from
our solvers to urge them again to an attentive and patient
review of every point in our text that may not have been
clear. Some of our examples, probably many of them,
WHITE.
have been understood by every reader; but the interest
of the retrograde theory as a whole so far transcends that
of any single analysis, however brilliant that it is well
· Self-mate in one. worth while going over every doubtful p~int until it has
been m~s~ered. Space requirements, and the difficulties
o! expla1n1ng a new subject have perhaps made us some-
times _obscure, and we rely on the reader's sympathetic
study to remedy our shortcomings.
 I P)$QPe.p.+.
. No. 95. At first sight it looks as.though there were a remarkable
cook by I P x S P e.p. +. The move is remarkable,
·indeed, but onlv as a try, for the retrogression : I Pb7-
. W. HUNDSDORFER.
b5, Ka4-a5 ; - 2 Pd7-d5 + bottles up· the square c~,
so the Q B can· never return thither. The true analysis
·• Tidskrift, Igo8. reads : I Pd7-d5, Ka-J,-aS ; 2 Pb6-b5 +, Ka3-a.q. ;
BI,ACK.
3 Pb5-b4+. .
No. 95A presents a lateral cover, I. Pe7-e5 protectmg
c7 from the Rf7, though,~he real_ ~eauty lies in th~ mat~e- ,
l matical "odd-and-even)(' opposition of the White Kmg:,eJ~.tA'.-14,
and the un~pturecCB1ack Kn~ght, :who m1:st re~ppear just at € j 1~-r
'the correct instant, or an imagmary situation deyelops. /',
The last moves were I Pe7-e5 ! Pfa XS (now follow
any number of oscillatiqns of the White King on c7, c8,
till at last must come) ; zS-c6 or d7, Bd6-b8 ; 3 Sb8-
c6, d7 +, Kc7-c8 ; etc. If we retracted only I Pe6--e5,
. the1e would follow PfzxS.; 2 Sc4-e3, Kc7-c8; 3 Pe7-e
6 +, and now the Black S'~ii'sl arrive at c6 at the wrong moment.
No. 95B covers h5 from the Qe8. The try I Bf7-g6,
Kh5-h6 ; z Rg6-:--gJ +, Kh6--,.h5 ; 3 Rg7-g? + fs e~actly
reminiscent of Troitzky's perpetual retrogression m_l\o. 64.
' I . •• , No. 95A.
- - - - - - - - - - - ~ , ..i:.
No. 95B. 1 •
4!;-t:JfJ:t A. Triotzky. T. R .. J?.awsor, ·Y'
,/.Yp·1. ,·• ,.. :',.~/. Original.
BLACK.
Ongmal. , \I
BLACK. V
,.
16
. m ~~
-,?~1· 21M

WHITE.

Mate in two.

'IVHI'rE. WHITE. JO
Mate in four. Mate in two.
.I. I PxPe.p. . I SPxP,e.p.+ .

fi rt. ct\ v11V)l1 ; ; yt/..il·~ ;,,t ...Z I95 La

~ Ji;{......d,t ~1i i-~¼ r..,"''l-h-·/1-,,v-;
·. cld '?~ C,, 0. T,;, C l { 31'·1l c,(~ )
 No. 96. ,· No. 96.
,
.

T. R. DAWSON. I PxPe.p.+.
In all of groups SI-95 only two reasons have been
Original. shown why the theme P moved last, either because all
BLACK.
j, other Black force was in retrograde stalemate, or because
such a move was the only way to release White from
impending stalemate. There remains a simple, though
unorthodox, method to avoid all the constructional
difficulties of these retrograde stalemates. This is to
have the theme P check on the 9-iagram. Unless this
check is to be imaginary, it is perfectly obvious 1that the
theme P must have moved last. (Compare Nos. 20-21).
The conventional side-square type is, of course, out of
the question, s~nce the White K cannot stand both upon
the side-square and in check.

N"o. 96A. No. 96B.

A.· \Vaterhouse. T. R. Dawson.
·I -- Jl,Janchester.Times, 20 Dec., 1913. Chess Amateur, May, 19q.
BLACK. Bi.ACK.

I
. :

WHITE.

Mate in two.

Mate in two. Mate in two.

I
1 Px P e.p.+ .. l RP_x Pe.p.+.

1 97
 No. 97.
No. 97.:
T. R. DAWSON.
I PxPe.p.+.
V. Gazette Times, 23 Nov., 1913.
Excepting the side-square type, therefore, all the
BI,ACK. en passant analytic themes may theoretically be illustrated
with ease and economy by this unorthodox short cut to
the truth! Even Troitzky's beautiful extension of the
side-square type, seen in No. 64, is handled quite simply
in No. 96. (1 Pd7-d5 +, Kc5-c4; 2 S-a4 + ). The
blocking of the passage square has been prettily .shown
in this fashion by Waterhouse, in No. 96A (1 .Pb7-b5 +,
Sb6-c8). Line covering also offers excellent opportunities
for economy, as in No. 96n, where the theme P covers the·
Black K against Rh7-h8. As to cross-checks, we never
obtained more than three, as in No. 65. No. 97 readily
establishes a record of no less than six checks, the passage
square not being crossed till White's third last move,
thus : I Pe7-e5 +, Pg3-g4 + ; 2 Se4-c5 +, Rb4-
b6 +; 3 Sb6-d7+, Bd7-h3+ !
Our whole analysis of the en passant keys had its starting
point in a splendid essay prepared in 1909 at the present
editor's request by W. Hundsdorfer. An ).mfortunate
mental affliction prevented its completion, artd it was
not till 1913 that T. R. Dawson took it up, and completely
\ rewrote and rearranged it, establishing the classification
\ WHI'.I'E. and preparing the analyses here adopted.
\\
i
'l'he editor's share in the work bas been further mini-
\ Mate in two. mised by the manifold assistance of Mi. Murray Marble,
the vigorous testing of Mr. E. L. Dashiell, and the copying
\fcJ; -1R~i -frt~,\,{.u.e of Mr. Malcolm Zim. All of which he acknowledges with
gratitude, as well as the. magnificent original contributions
of Herr Troitzky. The proofs have been read by T. R.
Dawson. ·

198 199

t
 (fi=.:_~r~~,:~·1
No. 98. ,1
~v
We have_ already encountered,· in Nos. IO-II, some
·T. R. DAWSON. manifestations' "-of that curious phenomenon we termed
·Original.
r ~ rti 1 A al¥ !:',> In those early groups arialysis revealed
p esence of some illegality, but did not determine
BLACK.
3 '
accurately what the illegality was. Each of several
plausible explanations, indeed, allowed us to put a <;li:fferent
1
correction and solution into effect. We now meet retro-
grade partial analysis in the domain of legality, instead of
in the domain of illegality. In each problem of the present
_group it can be determined analytically that a Black P
has just moved two squares, but it cannot be determined
exactly which P has done so. '.L'he problems are solved,
therefore, by capturing en passant whatever P has just
moved, seven sound legal en passant keys for these three
positions. ·
Ko. 98A. No. 98B.
C. G. VVatson. T. R. Dawson. /
1vlelbourne Leader, 25 Oct., 1913. 1vfelbourne Leader, 29 Aug., 19q.
BLACK.

WHITE.

Mate in two.

Mate in three. Mate in two.

· i: PX P e.p. accordingly. 1 PxP e.p: accordingly. ·

201
. ,,
 ~
...

...
No.-99. .'. j ·. ::.~
~"

i
N. HQEG. No. 99.
r
Wochenschach, 2· June, 1907.
Partial analysis may affect Castling, as well as the
en passant capture, both being conditioned on retrograde

... ~.,,~-~~
••••
BLACK.
t:;,. .. circumstances. Readers of the Loyd book will remember
"'
1■
his Nos. 64-5, versions of an idea of his dating back to
1859, in which. mate was possible in two distinct ways

■ ■ ~~ -~
•,1•■it••·
~,, ~l'I ■! rt,.,ai Dlt)
•
according to Black's previous play. Loyd enjoyed the
fun these analytics gave him with persons who claimed
that problems were not properly endings from actual
play (see Loyd book, p. 61) ; but he never troubled to
elaborate the possibilities of partial analysis, beyond
giving Black the choice on which side to Castle in the two
•
-_WJt■T■WJt■'
~'~( ~ ~ ~ ~ ~ ~ . problems just referred to. Loyd's mechanism was to set
all of Black's force except the Kand two R's in immediate
retrograde stalemate, revealing at once to the initiated
· that one or other of the R's must have moved. In No.
,.,,, . . . .7,'"'
i ■ II I ~ fJJ 1
100B the same alternative is much better disguised by
omitting the retrograde stalemate. It takes a careful

,~•~,itli'!: --
count of heads to discover that a White P has promoted
either at a8 or at h8, forcing one R to move and permitting
>
mate by 1 Qe5 or 1 Qe4 accordingly. A composer named
! ,.,.. Ss;.hos.t~Fk, of whom ·,ve have not otherwise ht!ard, was
'-., theear1iest (1874) to try to combi11e the Castling contin-
~
,,
~- gency with the en passant capture. His version was
WHl'.l'E.
/ unsound, and was followed by later sound examples by
; .! Oeffner (No. 100A) and J. Keeble, 1890. In No. 100A
Mate in three. ,;
,., . Black claims the privilege to Castle, and White retaliates
.·, by demonstrating the right to 1 P XP e.p. If both sides
surrender these weapons, as they must do mutually, then
mate follows by 1 K xP.

202 203
 No. 100.

No. 100. These simpler partial analytics __ pa~e before ~he com-
plexities of Nos. 99 and IOO. Hoeg s masterpiece was
composed before the re1;aissance of retrograde study -,/'
T. R. DAWSON".
narrated on p. I37, and is all the more noteworthy on ("'-
that account. 'If Black's last move was made by QR,
Original. I Sg6 solves. If by K R, I Sc6 solves. !f Black can
still Castle on both sides the last move was either I Pd7-
BLACK. d5 or I Pf7-f5, and e1; passant ~eys are legalised lo'.ccord-
ingly ; sur_ely the acme of ap. partial retrograde stalemates.
In No. mo. seven Black pieces have ~een captured, t~e
White P's requiring only six. Hence either ~me of Blacks
R P's may have died unpromoted, and \Vh1te can Castle
on that sioe. But if Black's last move was Pd7-d5 or
Pe7-e5, then a corresponding Black B ~ell ~m its home
square, both R P's promoted, and Castling ~s outlawed.
Fortunately r P x Q P e.p. or I P X KP e.p. is now legal
and .effective. Again, one problem and four keys. Re-
trograde contingencies can go no further, so we must
herewith conclude_ our analyses !

Ko. IOOA. ::-ro. ioon.

' A. Oeffner. T. R. Dawson.
Brentano's, 1882. Original.
BLACK. BLACK.

WHITE:

Mate in two.

Mate in three'. Mate in two.

205
 INDEX OF COMPOSERS.

Anonymous, 53.
Amelung, 55B.
Babson, 57.B.
Barry, I7, 38.
Batori, 9, 72, 75B.
Behal, 62.
Bernard, I3B.
Bertrand, I9B.
Bolus, I6, 4IA.
Chocholous, IIA.
· -· Collins, I9, 3I, 47-
Cook, 54-
Crake, I8.
Davis, I7A.
Dawson, IB, JB, I4, IJB, 2i, 22, 23, 23A, 24, 25, 25A-B,
26, 28, 34, 35, 35A-B, ·3( 37,. 37A-B, 42, 43, 43A-B,
44, 4,s, 45B, 46, 48, 49, 49A~B, 50, SI, 5IA-.B, 6m,
63.A-~; 65, 67B, JI, JIA-B, 73, 7}A~B, 74, 78, 8IA,
83A-B, 85, 87: 87A, 8.9, 93, 94, 95:i3, 96, 96n, 97, •98'.
9~B,,IOO, IOOB .. · .
Dusold, 27 A.
. , Eginton, ·3A, I6, 65B .
i
l Ewetzky, 4 ..
. r' ·,
 1~~/;~~!!~J;;'t,
;;,:;, ':' > , Hermann, 5B. ·
.,,:::,}{;\:' ~jo~~~/f ·, . <.. ' - • :
}./1 ~ Hoeg, ~21~~~, 47~,
Hume,· 7, 29A, · ,: :>,:Robertson,·. 9~. _, -· . · · ·I

·1'~;!~:t .
Hmidsdorf~r, 59B, 60, 63,, 65, 67, 67A, 68, 69A-B, 72, ?5, :>~f~Rowland: 8, q,7, 29~, 30.
75A, 79B, 80, 8m, 82, 83, 84, 86, 87B, 88, 90, 90B,
91__, 9lA-B, 9i, _95 ..
Jespersen,_ 15, I,5A-B, 2JB, 27B.
De Jong; IA.
· 'Steinweg, IJA.
Keeble, IIB.
Kennard, 6r. .
· ... Troitzky, 3m, 64, 68, 69, 70; 7?, J7B, 79, 79A, 8I, 83, 85B,
·'·'i·:.' -90A, . 9. 5A .
-..:·.. . . .
King Parks,' I.
'1on' Ungern-Sternberg, 77A'.
Kipping, 4m.
Kireewsley, II.
'W~nw~ght,_ ~9; 39A7B, 40, 6IA.
, 5; 65A, 96A.
_Klett, ·52,
· onig, 9i. . .
:'·-.;' .,.:.·,,. _,